From 7797f4956918f1631c52714034ee1690b8fd3fd3 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:20:41 +0300
Subject: [PATCH 01/82] test(api): assert parent HelmRelease Install/Upgrade
 Timeout >= 15m

Adds a failing unit test for convertApplicationToHelmRelease asserting
that Install.Timeout and Upgrade.Timeout are at least 15 minutes. The
default flux helm-controller timeout is too short to cover cold-start
Kamaji control-plane bootstrap (image pull + etcd bootstrap + apiserver
Ready + admin-kubeconfig Secret generation) and causes install
remediation loops.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../apps/application/rest_timeout_test.go     | 53 +++++++++++++++++++
 1 file changed, 53 insertions(+)
 create mode 100644 pkg/registry/apps/application/rest_timeout_test.go

diff --git a/pkg/registry/apps/application/rest_timeout_test.go b/pkg/registry/apps/application/rest_timeout_test.go
new file mode 100644
index 00000000..8ae13183
--- /dev/null
+++ b/pkg/registry/apps/application/rest_timeout_test.go
@@ -0,0 +1,53 @@
+package application
+
+import (
+	"testing"
+	"time"
+
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+
+	appsv1alpha1 "github.com/cozystack/cozystack/pkg/apis/apps/v1alpha1"
+	"github.com/cozystack/cozystack/pkg/config"
+)
+
+func TestConvertApplicationToHelmRelease_SetsInstallAndUpgradeTimeout(t *testing.T) {
+	r := &REST{
+		releaseConfig: config.ReleaseConfig{
+			Prefix: "kubernetes-",
+			ChartRef: config.ChartRefConfig{
+				Kind:      "HelmChart",
+				Name:      "kubernetes",
+				Namespace: "cozy-system",
+			},
+		},
+	}
+
+	app := &appsv1alpha1.Application{
+		ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
+	}
+
+	hr, err := r.convertApplicationToHelmRelease(app)
+	if err != nil {
+		t.Fatalf("convertApplicationToHelmRelease returned error: %v", err)
+	}
+
+	if hr.Spec.Install == nil {
+		t.Fatal("Spec.Install must not be nil")
+	}
+	if hr.Spec.Install.Timeout == nil {
+		t.Fatal("Spec.Install.Timeout must be set to cover async admin-kubeconfig provisioning")
+	}
+	if hr.Spec.Install.Timeout.Duration < 15*time.Minute {
+		t.Errorf("Spec.Install.Timeout must be >= 15m (cold bootstrap budget), got %v", hr.Spec.Install.Timeout.Duration)
+	}
+
+	if hr.Spec.Upgrade == nil {
+		t.Fatal("Spec.Upgrade must not be nil")
+	}
+	if hr.Spec.Upgrade.Timeout == nil {
+		t.Fatal("Spec.Upgrade.Timeout must be set to cover async admin-kubeconfig provisioning")
+	}
+	if hr.Spec.Upgrade.Timeout.Duration < 15*time.Minute {
+		t.Errorf("Spec.Upgrade.Timeout must be >= 15m (cold bootstrap budget), got %v", hr.Spec.Upgrade.Timeout.Duration)
+	}
+}

From e4f279f8e2cb08e271130b4a5193d9e0e9d9d3a7 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:21:15 +0300
Subject: [PATCH 02/82] fix(api): set 15m Install/Upgrade Timeout for parent
 HelmRelease

Parent HelmRelease created by cozystack-api for Kubernetes tenants
contains CAPI/Kamaji resources (Cluster, KamajiControlPlane,
MachineDeployment) that asynchronously provision the
*-admin-kubeconfig Secret. Three Deployments in the same chart
(cluster-autoscaler, kccm, kcsi-controller) mount that Secret
directly, so the helm-wait cannot complete until control-plane
bootstrap finishes.

Default flux helm-controller timeout is too short for a cold-node
first-tenant bootstrap (image pull + etcd bootstrap + apiserver Ready
+ admin-kubeconfig generation routinely exceed it). On timeout,
install.remediation triggers uninstall, which removes the Cluster CR
and restarts the cycle indefinitely.

Bumping Install.Timeout and Upgrade.Timeout to 15m gives realistic
bootstrap headroom while remaining bounded.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 pkg/registry/apps/application/rest.go | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/pkg/registry/apps/application/rest.go b/pkg/registry/apps/application/rest.go
index 0728ea13..77d1d458 100644
--- a/pkg/registry/apps/application/rest.go
+++ b/pkg/registry/apps/application/rest.go
@@ -1509,11 +1509,13 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (*
 			},
 			Interval: metav1.Duration{Duration: 5 * time.Minute},
 			Install: &helmv2.Install{
+				Timeout: &metav1.Duration{Duration: 15 * time.Minute},
 				Remediation: &helmv2.InstallRemediation{
 					Retries: -1,
 				},
 			},
 			Upgrade: &helmv2.Upgrade{
+				Timeout: &metav1.Duration{Duration: 15 * time.Minute},
 				Remediation: &helmv2.UpgradeRemediation{
 					Retries: -1,
 				},

From 3e26234a1c98f138542471a7f4395532e15c7b46 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:24:20 +0300
Subject: [PATCH 03/82] test(kubernetes): assert admin-kubeconfig wait pattern
 and etcd guard

Adds failing helm unittest suite for packages/apps/kubernetes covering:

- cluster-autoscaler, kccm, and csi controller Deployments mount the
  admin-kubeconfig Secret with optional: true
- each of those Deployments has a wait-for-kubeconfig initContainer
  that mounts the same kubeconfig path
- cluster.yaml renders a helm fail with a descriptive message when the
  tenant has no etcd DataStore (empty _namespace.etcd)

Also wires up a test target in the chart Makefile so helm-unit-tests.sh
picks it up.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/Makefile             |  3 +
 .../tests/admin_kubeconfig_wait_test.yaml     | 85 +++++++++++++++++++
 packages/apps/kubernetes/tests/values-ci.yaml |  9 ++
 3 files changed, 97 insertions(+)
 create mode 100644 packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
 create mode 100644 packages/apps/kubernetes/tests/values-ci.yaml

diff --git a/packages/apps/kubernetes/Makefile b/packages/apps/kubernetes/Makefile
index 01cf736d..0f9e6d57 100644
--- a/packages/apps/kubernetes/Makefile
+++ b/packages/apps/kubernetes/Makefile
@@ -4,6 +4,9 @@ KUBERNETES_PKG_TAG = $(shell awk '$$1 == "version:" {print $$2}' Chart.yaml)
 include ../../../hack/common-envs.mk
 include ../../../hack/package.mk
 
+test:
+	helm unittest .
+
 generate:
 	cozyvalues-gen -m 'kubernetes' -v values.yaml -s values.schema.json -r README.md -g ../../../api/apps/v1alpha1/kubernetes/types.go
 	../../../hack/update-crd.sh
diff --git a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
new file mode 100644
index 00000000..e2d9f359
--- /dev/null
+++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
@@ -0,0 +1,85 @@
+suite: admin-kubeconfig wait guards
+
+release:
+  name: test
+  namespace: tenant-root
+
+values:
+  - values-ci.yaml
+
+tests:
+  - it: cluster-autoscaler mounts admin-kubeconfig as optional
+    template: templates/cluster-autoscaler/deployment.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
+    asserts:
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.secretName
+          value: test-admin-kubeconfig
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.optional
+          value: true
+
+  - it: cluster-autoscaler waits for admin-kubeconfig via initContainer
+    template: templates/cluster-autoscaler/deployment.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
+    asserts:
+      - equal:
+          path: spec.template.spec.initContainers[0].name
+          value: wait-for-kubeconfig
+      - contains:
+          path: spec.template.spec.initContainers[0].volumeMounts
+          content:
+            name: kubeconfig
+            mountPath: /etc/kubernetes/kubeconfig
+            readOnly: true
+
+  - it: kccm mounts admin-kubeconfig as optional
+    template: templates/kccm/manager.yaml
+    asserts:
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.secretName
+          value: test-admin-kubeconfig
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.optional
+          value: true
+
+  - it: kccm waits for admin-kubeconfig via initContainer
+    template: templates/kccm/manager.yaml
+    asserts:
+      - equal:
+          path: spec.template.spec.initContainers[0].name
+          value: wait-for-kubeconfig
+
+  - it: csi controller mounts admin-kubeconfig as optional
+    template: templates/csi/deploy.yaml
+    asserts:
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.secretName
+          value: test-admin-kubeconfig
+      - equal:
+          path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.optional
+          value: true
+
+  - it: csi controller waits for admin-kubeconfig via initContainer
+    template: templates/csi/deploy.yaml
+    asserts:
+      - equal:
+          path: spec.template.spec.initContainers[0].name
+          value: wait-for-kubeconfig
+
+  - it: cluster.yaml fails render when tenant has no etcd DataStore
+    template: templates/cluster.yaml
+    set:
+      _namespace:
+        etcd: ""
+        monitoring: ""
+        ingress: ""
+        seaweedfs: ""
+        host: ""
+    asserts:
+      - failedTemplate:
+          errorPattern: "requires a parent tenant with etcd enabled"
diff --git a/packages/apps/kubernetes/tests/values-ci.yaml b/packages/apps/kubernetes/tests/values-ci.yaml
new file mode 100644
index 00000000..13365e8c
--- /dev/null
+++ b/packages/apps/kubernetes/tests/values-ci.yaml
@@ -0,0 +1,9 @@
+_namespace:
+  etcd: tenant-root
+  monitoring: ""
+  ingress: ""
+  seaweedfs: ""
+  host: ""
+_cluster:
+  cluster-domain: cozy.local
+nodeGroups: null

From ca33cc4e3c4b032293fceec2b7d3c1eb13a241bb Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:25:42 +0300
Subject: [PATCH 04/82] fix(kubernetes): wait for admin-kubeconfig before
 starting CP-side pods

Three Deployments in the Kubernetes app chart mount the tenant
admin-kubeconfig Secret directly as a volume: cluster-autoscaler,
kccm, and the kcsi controller. That Secret is provisioned
asynchronously by Kamaji after control-plane bootstrap, so on a fresh
install the pods used to hit FailedMount and the parent HelmRelease
ran out of its wait budget.

Mark the Secret volume optional and add a wait-for-kubeconfig
initContainer that polls the mounted path until the Secret appears.
Kubelet remounts the optional Secret within its sync period once
Kamaji publishes it, the init container exits, and the main container
starts cleanly. The Deployment becomes Available without the helm-wait
ever seeing a FailedMount.

Pins a busybox image for the init container via
images/busybox.tag (same format as the other pinned tags in this
chart).

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/images/busybox.tag     |  1 +
 .../cluster-autoscaler/deployment.yaml          | 17 +++++++++++++++++
 .../apps/kubernetes/templates/csi/deploy.yaml   | 17 +++++++++++++++++
 .../apps/kubernetes/templates/kccm/manager.yaml | 17 +++++++++++++++++
 4 files changed, 52 insertions(+)
 create mode 100644 packages/apps/kubernetes/images/busybox.tag

diff --git a/packages/apps/kubernetes/images/busybox.tag b/packages/apps/kubernetes/images/busybox.tag
new file mode 100644
index 00000000..e358c12e
--- /dev/null
+++ b/packages/apps/kubernetes/images/busybox.tag
@@ -0,0 +1 @@
+busybox:1.37.0@sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e
diff --git a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
index a00e0155..47c22f76 100644
--- a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
+++ b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
@@ -23,6 +23,22 @@ spec:
         - key: node-role.kubernetes.io/control-plane
           operator: Exists
           effect: "NoSchedule"
+      initContainers:
+      - name: wait-for-kubeconfig
+        image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
+        command:
+        - sh
+        - -c
+        - |
+          set -eu
+          until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+            echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
+            sleep 5
+          done
+        volumeMounts:
+        - name: kubeconfig
+          mountPath: /etc/kubernetes/kubeconfig
+          readOnly: true
       containers:
       - image: "{{ $.Files.Get "images/cluster-autoscaler.tag" | trim }}"
         name: cluster-autoscaler
@@ -56,6 +72,7 @@ spec:
         name: cloud-config
       - secret:
           secretName: {{ .Release.Name }}-admin-kubeconfig
+          optional: true
         name: kubeconfig
       serviceAccountName: {{ .Release.Name }}-cluster-autoscaler
       terminationGracePeriodSeconds: 10
diff --git a/packages/apps/kubernetes/templates/csi/deploy.yaml b/packages/apps/kubernetes/templates/csi/deploy.yaml
index 938b6d67..2979c75c 100644
--- a/packages/apps/kubernetes/templates/csi/deploy.yaml
+++ b/packages/apps/kubernetes/templates/csi/deploy.yaml
@@ -24,6 +24,22 @@ spec:
         - key: node-role.kubernetes.io/control-plane
           operator: Exists
           effect: "NoSchedule"
+      initContainers:
+        - name: wait-for-kubeconfig
+          image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
+          command:
+          - sh
+          - -c
+          - |
+            set -eu
+            until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+              echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
+              sleep 5
+            done
+          volumeMounts:
+          - name: kubeconfig
+            mountPath: /etc/kubernetes/kubeconfig
+            readOnly: true
       containers:
         - name: csi-driver
           imagePullPolicy: Always
@@ -234,4 +250,5 @@ spec:
           emptyDir: {}
         - secret:
             secretName: {{ .Release.Name }}-admin-kubeconfig
+            optional: true
           name: kubeconfig
diff --git a/packages/apps/kubernetes/templates/kccm/manager.yaml b/packages/apps/kubernetes/templates/kccm/manager.yaml
index 81426d4e..5c7f15a4 100644
--- a/packages/apps/kubernetes/templates/kccm/manager.yaml
+++ b/packages/apps/kubernetes/templates/kccm/manager.yaml
@@ -22,6 +22,22 @@ spec:
         - key: node-role.kubernetes.io/control-plane
           operator: Exists
           effect: "NoSchedule"
+      initContainers:
+      - name: wait-for-kubeconfig
+        image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
+        command:
+        - sh
+        - -c
+        - |
+          set -eu
+          until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+            echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
+            sleep 5
+          done
+        volumeMounts:
+        - name: kubeconfig
+          mountPath: /etc/kubernetes/kubeconfig
+          readOnly: true
       containers:
       - name: kubevirt-cloud-controller-manager
         args:
@@ -55,5 +71,6 @@ spec:
         name: cloud-config
       - secret:
           secretName: {{ .Release.Name }}-admin-kubeconfig
+          optional: true
         name: kubeconfig
       serviceAccountName: {{ .Release.Name }}-kccm

From cac514b60fc00717df37780d2183e1702b07e9c6 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:26:09 +0300
Subject: [PATCH 05/82] fix(kubernetes): fail fast when tenant has no etcd
 DataStore

When a Kubernetes tenant is created without a parent tenant that has
etcd enabled, .Values._namespace.etcd is empty and the rendered
KamajiControlPlane spec carries an empty dataStoreName. The Kamaji
admission webhook then rejects every TenantControlPlane create with
"tenant-root DataStore does not exist" and the control plane never
comes up.

Add a helm template-level guard that fails rendering with a
descriptive, actionable error message before the HelmRelease even
reaches the webhook. This also closes the narrow race where a
Kubernetes HelmRelease reconciles before the etcd HelmRelease has
created the DataStore CR - flux retries on its interval and picks up
the DataStore once it appears.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/templates/cluster.yaml | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/packages/apps/kubernetes/templates/cluster.yaml b/packages/apps/kubernetes/templates/cluster.yaml
index 10d6fd80..99249c1a 100644
--- a/packages/apps/kubernetes/templates/cluster.yaml
+++ b/packages/apps/kubernetes/templates/cluster.yaml
@@ -1,4 +1,7 @@
 {{- $etcd := .Values._namespace.etcd }}
+{{- if not $etcd }}
+{{- fail "Kubernetes tenant requires a parent tenant with etcd enabled: set .Values.etcd=true on tenant-root (or any ancestor tenant), wait for the etcd HelmRelease to reconcile (DataStore CR appears in the tenant namespace), then retry." }}
+{{- end }}
 {{- $ingress := .Values._namespace.ingress }}
 {{- $host := .Values._namespace.host }}
 {{- $kubevirtmachinetemplateNames := list }}

From a7d994365d067dee5d9546a9b95bb555b206a6af Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:26:47 +0300
Subject: [PATCH 06/82] test(kubernetes): assert parent HelmRelease did not
 remediate in e2e

Before cleanup, inspect the parent HelmRelease installFailures and
upgradeFailures counters. A non-zero value means flux helm-controller
hit its wait timeout, ran install/upgrade remediation (uninstall),
and re-installed - the exact race condition this PR closes. Fail the
bats test in that case so the signal surfaces in CI instead of being
masked by a green retry.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/e2e-apps/run-kubernetes.sh | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/hack/e2e-apps/run-kubernetes.sh b/hack/e2e-apps/run-kubernetes.sh
index be6dcd6f..f05ec370 100644
--- a/hack/e2e-apps/run-kubernetes.sh
+++ b/hack/e2e-apps/run-kubernetes.sh
@@ -320,6 +320,19 @@ EOF
     done
     kubectl wait hr kubernetes-${test_name}-ingress-nginx -n tenant-test --timeout=5m --for=condition=ready
 
+  # Guard: parent HelmRelease must not have entered an install/upgrade remediation cycle.
+  # A non-zero installFailures/upgradeFailures indicates the helm-wait budget expired while
+  # admin-kubeconfig was still being provisioned, which would trigger uninstall remediation
+  # and churn the Cluster CR.
+  install_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.installFailures}')
+  upgrade_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.upgradeFailures}')
+  if [ "${install_failures:-0}" != "0" ] && [ -n "${install_failures}" ] || \
+     [ "${upgrade_failures:-0}" != "0" ] && [ -n "${upgrade_failures}" ]; then
+    echo "Parent HelmRelease entered remediation cycle: installFailures=${install_failures:-0}, upgradeFailures=${upgrade_failures:-0}" >&2
+    kubectl -n tenant-test describe hr "kubernetes-${test_name}" >&2
+    exit 1
+  fi
+
   # Clean up
   pkill -f "port-forward.*${port}:" 2>/dev/null || true
   rm -f "tenantkubeconfig-${test_name}"

From f87834a3edba32945f22e8d419a9592e5303b34b Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:48:47 +0300
Subject: [PATCH 07/82] fix(hack): group e2e remediation guard conditions
 correctly

Shell && and || have equal precedence and left-to-right associativity,
so the previous guard parsed as (((A && B) || C) && D) and silently
passed on the canonical failure mode: install_failures=1 with an empty
upgrade_failures.

Extract the check into helmrelease_has_remediation_cycle() in a
dedicated helper sourced from run-kubernetes.sh, and add unit tests
under hack/remediation-guard.bats that pin the expected behavior for
every combination of empty, zero, and positive counter values.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/e2e-apps/remediation-guard.sh | 24 +++++++++
 hack/e2e-apps/run-kubernetes.sh    |  5 +-
 hack/remediation-guard.bats        | 87 ++++++++++++++++++++++++++++++
 3 files changed, 114 insertions(+), 2 deletions(-)
 create mode 100644 hack/e2e-apps/remediation-guard.sh
 create mode 100644 hack/remediation-guard.bats

diff --git a/hack/e2e-apps/remediation-guard.sh b/hack/e2e-apps/remediation-guard.sh
new file mode 100644
index 00000000..6da69f13
--- /dev/null
+++ b/hack/e2e-apps/remediation-guard.sh
@@ -0,0 +1,24 @@
+# Helpers for asserting that a Flux HelmRelease did not fall into an
+# install/upgrade remediation cycle during an e2e run.
+#
+# A non-zero installFailures/upgradeFailures counter means flux
+# helm-controller hit its wait timeout, ran remediation (uninstall),
+# and re-installed. That is exactly the race this guard is meant to
+# catch, so the function returns success (0) when a cycle is detected
+# and failure (1) otherwise.
+#
+# Both arguments may be empty strings, the literal "0", or a positive
+# integer. Shell's && and || have equal precedence with left-to-right
+# associativity, so each half of the disjunction is grouped explicitly
+# to avoid (A && B) || C && D parsing that masks the common
+# install_failures=1, upgrade_failures="" case.
+
+helmrelease_has_remediation_cycle() {
+    install_failures="$1"
+    upgrade_failures="$2"
+    if { [ -n "${install_failures}" ] && [ "${install_failures}" != "0" ]; } || \
+       { [ -n "${upgrade_failures}" ] && [ "${upgrade_failures}" != "0" ]; }; then
+        return 0
+    fi
+    return 1
+}
diff --git a/hack/e2e-apps/run-kubernetes.sh b/hack/e2e-apps/run-kubernetes.sh
index f05ec370..811f42a7 100644
--- a/hack/e2e-apps/run-kubernetes.sh
+++ b/hack/e2e-apps/run-kubernetes.sh
@@ -1,3 +1,5 @@
+. hack/e2e-apps/remediation-guard.sh
+
 run_kubernetes_test() {
     local version_expr="$1"
     local test_name="$2"
@@ -326,8 +328,7 @@ EOF
   # and churn the Cluster CR.
   install_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.installFailures}')
   upgrade_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.upgradeFailures}')
-  if [ "${install_failures:-0}" != "0" ] && [ -n "${install_failures}" ] || \
-     [ "${upgrade_failures:-0}" != "0" ] && [ -n "${upgrade_failures}" ]; then
+  if helmrelease_has_remediation_cycle "${install_failures}" "${upgrade_failures}"; then
     echo "Parent HelmRelease entered remediation cycle: installFailures=${install_failures:-0}, upgradeFailures=${upgrade_failures:-0}" >&2
     kubectl -n tenant-test describe hr "kubernetes-${test_name}" >&2
     exit 1
diff --git a/hack/remediation-guard.bats b/hack/remediation-guard.bats
new file mode 100644
index 00000000..64aa83af
--- /dev/null
+++ b/hack/remediation-guard.bats
@@ -0,0 +1,87 @@
+#!/usr/bin/env bats
+# -----------------------------------------------------------------------------
+# Unit tests for hack/e2e-apps/remediation-guard.sh
+#
+# helmrelease_has_remediation_cycle is consumed from e2e tests to assert that
+# the parent HelmRelease did not hit flux helm-controller's wait timeout and
+# enter uninstall remediation. The function accepts two arguments (values of
+# .status.installFailures and .status.upgradeFailures) and returns 0 when a
+# remediation cycle is detected, 1 otherwise.
+#
+# Each argument can be empty (controller never populated the field), "0"
+# (populated but never failed), or a positive integer. Shell's && and ||
+# have equal precedence with left-to-right associativity, which used to
+# break this check on the most common failure mode - install_failures=1
+# and upgrade_failures="". These tests pin the correct behavior.
+#
+# cozytest.sh's awk parser recognizes only @test blocks and a bare `}` on
+# its own line; there is no bats `run` or `$status`. Assertions are
+# expressed as direct shell tests that exit non-zero on failure.
+#
+# Run with: hack/cozytest.sh hack/remediation-guard.bats
+# -----------------------------------------------------------------------------
+
+@test "no counters set returns not-detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "" "" || rc=$?
+    [ "$rc" -eq 1 ]
+}
+
+@test "both counters zero returns not-detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "0" "0" || rc=$?
+    [ "$rc" -eq 1 ]
+}
+
+@test "install zero upgrade empty returns not-detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "0" "" || rc=$?
+    [ "$rc" -eq 1 ]
+}
+
+@test "install empty upgrade zero returns not-detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "" "0" || rc=$?
+    [ "$rc" -eq 1 ]
+}
+
+@test "install one upgrade empty returns detected" {
+    # Canonical race: first install exceeded helm-wait, remediation fired,
+    # no upgrade has happened yet.
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "1" "" || rc=$?
+    [ "$rc" -eq 0 ]
+}
+
+@test "install empty upgrade one returns detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "" "1" || rc=$?
+    [ "$rc" -eq 0 ]
+}
+
+@test "install two upgrade zero returns detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "2" "0" || rc=$?
+    [ "$rc" -eq 0 ]
+}
+
+@test "install zero upgrade two returns detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "0" "2" || rc=$?
+    [ "$rc" -eq 0 ]
+}
+
+@test "both counters positive returns detected" {
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "3" "5" || rc=$?
+    [ "$rc" -eq 0 ]
+}

From 73b80bfb940ec24459a826b32540f676a00bca36 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:51:32 +0300
Subject: [PATCH 08/82] fix(api): scope 15m helm wait budget to Kubernetes
 Application kind

The previous change set Install.Timeout and Upgrade.Timeout to 15m
on every Application's parent HelmRelease, but the admin-kubeconfig
race documented in #2412 is specific to the Kubernetes kind: only its
parent chart creates CAPI/Kamaji resources whose admin-kubeconfig
Secret is asynchronously provisioned and mounted by Deployments in
the same chart. Other kinds (Qdrant, MongoDB, Postgres, ...) have no
such race and should not have their failed installs linger three times
longer before flux triggers remediation.

Gate the timeout on r.kindName == "Kubernetes". Rewrite
rest_timeout_test.go to cover both sides: Kubernetes must get a
>= 15m timeout, other kinds must keep the flux defaults. Both tests
also pin Install/Upgrade Remediation.Retries == -1 so a future edit
that removes unbounded remediation would show up as a red test.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 pkg/registry/apps/application/rest.go         | 16 ++++-
 .../apps/application/rest_timeout_test.go     | 58 ++++++++++++++++---
 2 files changed, 64 insertions(+), 10 deletions(-)

diff --git a/pkg/registry/apps/application/rest.go b/pkg/registry/apps/application/rest.go
index 77d1d458..9b92d2c0 100644
--- a/pkg/registry/apps/application/rest.go
+++ b/pkg/registry/apps/application/rest.go
@@ -1509,13 +1509,11 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (*
 			},
 			Interval: metav1.Duration{Duration: 5 * time.Minute},
 			Install: &helmv2.Install{
-				Timeout: &metav1.Duration{Duration: 15 * time.Minute},
 				Remediation: &helmv2.InstallRemediation{
 					Retries: -1,
 				},
 			},
 			Upgrade: &helmv2.Upgrade{
-				Timeout: &metav1.Duration{Duration: 15 * time.Minute},
 				Remediation: &helmv2.UpgradeRemediation{
 					Retries: -1,
 				},
@@ -1530,6 +1528,20 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (*
 		},
 	}
 
+	// The Kubernetes Application's parent chart creates CAPI/Kamaji resources
+	// whose admin-kubeconfig Secret is provisioned asynchronously and mounted
+	// by Deployments in the same chart. On a cold node, Kamaji control-plane
+	// bootstrap routinely exceeds flux helm-controller's default wait budget,
+	// so install remediation loops uninstall the Cluster CR and churn. Extend
+	// the wait budget to 15m for Kubernetes only - other Application kinds
+	// without this race keep flux defaults, so their failed installs do not
+	// linger unnecessarily before remediation fires.
+	if r.kindName == "Kubernetes" {
+		timeout := metav1.Duration{Duration: 15 * time.Minute}
+		helmRelease.Spec.Install.Timeout = &timeout
+		helmRelease.Spec.Upgrade.Timeout = &timeout
+	}
+
 	return helmRelease, nil
 }
 
diff --git a/pkg/registry/apps/application/rest_timeout_test.go b/pkg/registry/apps/application/rest_timeout_test.go
index 8ae13183..fa6d5427 100644
--- a/pkg/registry/apps/application/rest_timeout_test.go
+++ b/pkg/registry/apps/application/rest_timeout_test.go
@@ -10,18 +10,22 @@ import (
 	"github.com/cozystack/cozystack/pkg/config"
 )
 
-func TestConvertApplicationToHelmRelease_SetsInstallAndUpgradeTimeout(t *testing.T) {
-	r := &REST{
+func newRESTForKind(kind, prefix string) *REST {
+	return &REST{
+		kindName: kind,
 		releaseConfig: config.ReleaseConfig{
-			Prefix: "kubernetes-",
+			Prefix: prefix,
 			ChartRef: config.ChartRefConfig{
 				Kind:      "HelmChart",
-				Name:      "kubernetes",
+				Name:      "x",
 				Namespace: "cozy-system",
 			},
 		},
 	}
+}
 
+func TestConvertApplicationToHelmRelease_KubernetesKindGetsLongTimeout(t *testing.T) {
+	r := newRESTForKind("Kubernetes", "kubernetes-")
 	app := &appsv1alpha1.Application{
 		ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
 	}
@@ -35,19 +39,57 @@ func TestConvertApplicationToHelmRelease_SetsInstallAndUpgradeTimeout(t *testing
 		t.Fatal("Spec.Install must not be nil")
 	}
 	if hr.Spec.Install.Timeout == nil {
-		t.Fatal("Spec.Install.Timeout must be set to cover async admin-kubeconfig provisioning")
+		t.Fatal("Spec.Install.Timeout must be set for Kubernetes kind")
 	}
 	if hr.Spec.Install.Timeout.Duration < 15*time.Minute {
-		t.Errorf("Spec.Install.Timeout must be >= 15m (cold bootstrap budget), got %v", hr.Spec.Install.Timeout.Duration)
+		t.Errorf("Spec.Install.Timeout must be >= 15m for Kubernetes, got %v", hr.Spec.Install.Timeout.Duration)
 	}
 
 	if hr.Spec.Upgrade == nil {
 		t.Fatal("Spec.Upgrade must not be nil")
 	}
 	if hr.Spec.Upgrade.Timeout == nil {
-		t.Fatal("Spec.Upgrade.Timeout must be set to cover async admin-kubeconfig provisioning")
+		t.Fatal("Spec.Upgrade.Timeout must be set for Kubernetes kind")
 	}
 	if hr.Spec.Upgrade.Timeout.Duration < 15*time.Minute {
-		t.Errorf("Spec.Upgrade.Timeout must be >= 15m (cold bootstrap budget), got %v", hr.Spec.Upgrade.Timeout.Duration)
+		t.Errorf("Spec.Upgrade.Timeout must be >= 15m for Kubernetes, got %v", hr.Spec.Upgrade.Timeout.Duration)
+	}
+
+	if hr.Spec.Install.Remediation == nil || hr.Spec.Install.Remediation.Retries != -1 {
+		t.Errorf("Spec.Install.Remediation.Retries must remain -1, got %+v", hr.Spec.Install.Remediation)
+	}
+	if hr.Spec.Upgrade.Remediation == nil || hr.Spec.Upgrade.Remediation.Retries != -1 {
+		t.Errorf("Spec.Upgrade.Remediation.Retries must remain -1, got %+v", hr.Spec.Upgrade.Remediation)
+	}
+}
+
+func TestConvertApplicationToHelmRelease_NonKubernetesKindKeepsFluxDefaults(t *testing.T) {
+	// For Applications whose parent chart has no admin-kubeconfig race
+	// (Qdrant, MongoDB, Postgres, etc.), do NOT extend the helm-wait
+	// budget - otherwise failed installs would block three times as long
+	// before Flux starts remediating.
+	r := newRESTForKind("Qdrant", "qdrant-")
+	app := &appsv1alpha1.Application{
+		ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
+	}
+
+	hr, err := r.convertApplicationToHelmRelease(app)
+	if err != nil {
+		t.Fatalf("convertApplicationToHelmRelease returned error: %v", err)
+	}
+
+	if hr.Spec.Install != nil && hr.Spec.Install.Timeout != nil {
+		t.Errorf("Spec.Install.Timeout must be unset for non-Kubernetes kinds, got %v", hr.Spec.Install.Timeout.Duration)
+	}
+	if hr.Spec.Upgrade != nil && hr.Spec.Upgrade.Timeout != nil {
+		t.Errorf("Spec.Upgrade.Timeout must be unset for non-Kubernetes kinds, got %v", hr.Spec.Upgrade.Timeout.Duration)
+	}
+
+	// But remediation must still be -1 across the board.
+	if hr.Spec.Install.Remediation == nil || hr.Spec.Install.Remediation.Retries != -1 {
+		t.Errorf("Spec.Install.Remediation.Retries must remain -1, got %+v", hr.Spec.Install.Remediation)
+	}
+	if hr.Spec.Upgrade.Remediation == nil || hr.Spec.Upgrade.Remediation.Retries != -1 {
+		t.Errorf("Spec.Upgrade.Remediation.Retries must remain -1, got %+v", hr.Spec.Upgrade.Remediation)
 	}
 }

From 03606091df5672b6e857057a349875d534990f3a Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:53:17 +0300
Subject: [PATCH 09/82] chore(kubernetes): align busybox image with project
 convention

Adds images/busybox/Dockerfile and an image-busybox Makefile target
that mirror the same pattern as the rest of this chart's images (the
Dockerfile pins the upstream busybox by digest; the Makefile target
builds and tags for ghcr.io/cozystack/cozystack/busybox the same way
cluster-autoscaler et al. are handled). Also wires it into the
umbrella image target so 'make image' rebuilds everything.

Until the first release build runs image-busybox and rewrites the
.tag to point at ghcr.io, the .tag keeps a fully-qualified
docker.io/library/busybox:1.37.0@sha256:... reference so pods do not
silently resolve the short name via the default registry and pulls
remain immutable by digest. The release workflow overwrites this
file the same way it does for the other images.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/Makefile                  | 14 +++++++++++++-
 packages/apps/kubernetes/images/busybox.tag        |  2 +-
 packages/apps/kubernetes/images/busybox/Dockerfile |  1 +
 3 files changed, 15 insertions(+), 2 deletions(-)
 create mode 100644 packages/apps/kubernetes/images/busybox/Dockerfile

diff --git a/packages/apps/kubernetes/Makefile b/packages/apps/kubernetes/Makefile
index 0f9e6d57..eb23c2ef 100644
--- a/packages/apps/kubernetes/Makefile
+++ b/packages/apps/kubernetes/Makefile
@@ -15,7 +15,7 @@ update:
 	hack/update-versions.sh
 	make generate
 
-image: image-ubuntu-container-disk image-kubevirt-cloud-provider image-kubevirt-csi-driver image-cluster-autoscaler
+image: image-ubuntu-container-disk image-kubevirt-cloud-provider image-kubevirt-csi-driver image-cluster-autoscaler image-busybox
 
 image-ubuntu-container-disk:
 	$(foreach ver,$(KUBERNETES_VERSIONS), \
@@ -70,3 +70,15 @@ image-cluster-autoscaler:
 	echo "$(REGISTRY)/cluster-autoscaler:$(call settag,$(KUBERNETES_PKG_TAG))@$$(yq e '."containerimage.digest"' images/cluster-autoscaler.json -o json -r)" \
 		> images/cluster-autoscaler.tag
 	rm -f images/cluster-autoscaler.json
+
+image-busybox:
+	docker buildx build images/busybox \
+		--tag $(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG)) \
+		--tag $(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG)-$(TAG)) \
+		--cache-from type=registry,ref=$(REGISTRY)/busybox:latest \
+		--cache-to type=inline \
+		--metadata-file images/busybox.json \
+		$(BUILDX_ARGS)
+	echo "$(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG))@$$(yq e '."containerimage.digest"' images/busybox.json -o json -r)" \
+		> images/busybox.tag
+	rm -f images/busybox.json
diff --git a/packages/apps/kubernetes/images/busybox.tag b/packages/apps/kubernetes/images/busybox.tag
index e358c12e..39de220a 100644
--- a/packages/apps/kubernetes/images/busybox.tag
+++ b/packages/apps/kubernetes/images/busybox.tag
@@ -1 +1 @@
-busybox:1.37.0@sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e
+docker.io/library/busybox:1.37.0@sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e
diff --git a/packages/apps/kubernetes/images/busybox/Dockerfile b/packages/apps/kubernetes/images/busybox/Dockerfile
new file mode 100644
index 00000000..0c57fd39
--- /dev/null
+++ b/packages/apps/kubernetes/images/busybox/Dockerfile
@@ -0,0 +1 @@
+FROM docker.io/library/busybox:1.37.0@sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e

From 97696b2b036be79ecafd8e96401dc2af3bc39d32 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:54:10 +0300
Subject: [PATCH 10/82] test(kubernetes): add positive cluster render test and
 pin document kind

Adds an assertion that cluster.yaml renders successfully when
_namespace.etcd is set and produces a KamajiControlPlane whose
dataStoreName equals the tenant's etcd DataStore name. Without this
positive case a future edit that inverts or removes the existing etcd
guard would pass the suite as long as the negative case still fails.

Also adds documentSelector: kind=Deployment to the kccm and csi
controller assertions so the jsonpath filter operates on a single
document, matching the cluster-autoscaler case and removing
reliance on helm-unittest filter-vs-single-value coercion.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../tests/admin_kubeconfig_wait_test.yaml     | 34 +++++++++++++++++++
 1 file changed, 34 insertions(+)

diff --git a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
index e2d9f359..3b2b1d72 100644
--- a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
+++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
@@ -39,6 +39,9 @@ tests:
 
   - it: kccm mounts admin-kubeconfig as optional
     template: templates/kccm/manager.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
     asserts:
       - equal:
           path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.secretName
@@ -49,13 +52,25 @@ tests:
 
   - it: kccm waits for admin-kubeconfig via initContainer
     template: templates/kccm/manager.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
     asserts:
       - equal:
           path: spec.template.spec.initContainers[0].name
           value: wait-for-kubeconfig
+      - contains:
+          path: spec.template.spec.initContainers[0].volumeMounts
+          content:
+            name: kubeconfig
+            mountPath: /etc/kubernetes/kubeconfig
+            readOnly: true
 
   - it: csi controller mounts admin-kubeconfig as optional
     template: templates/csi/deploy.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
     asserts:
       - equal:
           path: spec.template.spec.volumes[?(@.name=="kubeconfig")].secret.secretName
@@ -66,10 +81,29 @@ tests:
 
   - it: csi controller waits for admin-kubeconfig via initContainer
     template: templates/csi/deploy.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
     asserts:
       - equal:
           path: spec.template.spec.initContainers[0].name
           value: wait-for-kubeconfig
+      - contains:
+          path: spec.template.spec.initContainers[0].volumeMounts
+          content:
+            name: kubeconfig
+            mountPath: /etc/kubernetes/kubeconfig
+            readOnly: true
+
+  - it: cluster.yaml renders and wires dataStoreName when tenant has etcd
+    template: templates/cluster.yaml
+    documentSelector:
+      path: kind
+      value: KamajiControlPlane
+    asserts:
+      - equal:
+          path: spec.dataStoreName
+          value: tenant-root
 
   - it: cluster.yaml fails render when tenant has no etcd DataStore
     template: templates/cluster.yaml

From 6afc0eb370f827c55f478045d1aff2a329b25100 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 20:55:17 +0300
Subject: [PATCH 11/82] fix(kubernetes): bound init wait and reword fail-fast
 message

Cap the wait-for-kubeconfig init container at 20m. If Kamaji genuinely
fails to produce the admin-kubeconfig Secret (misconfigured tenant,
etcd outage after the guard already passed, Kamaji crash-loop), the
pod now exits non-zero and goes CrashLoopBackOff so the failure is
visible in dashboards, instead of silently sleeping in Init forever
and leaving only the flux helm-wait timeout to surface the problem.

Reword the etcd DataStore guard to reference the parent Tenant
application's etcd flag (not .Values.etcd of the Kubernetes chart,
which is a different chart). Update the helm unittest errorPattern
accordingly.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../kubernetes/templates/cluster-autoscaler/deployment.yaml  | 5 +++++
 packages/apps/kubernetes/templates/cluster.yaml              | 2 +-
 packages/apps/kubernetes/templates/csi/deploy.yaml           | 5 +++++
 packages/apps/kubernetes/templates/kccm/manager.yaml         | 5 +++++
 .../apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml    | 2 +-
 5 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
index 47c22f76..277b6df8 100644
--- a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
+++ b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
@@ -31,7 +31,12 @@ spec:
         - -c
         - |
           set -eu
+          deadline=$(( $(date +%s) + 1200 ))
           until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+            if [ "$(date +%s)" -ge "$deadline" ]; then
+              echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
+              exit 1
+            fi
             echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
             sleep 5
           done
diff --git a/packages/apps/kubernetes/templates/cluster.yaml b/packages/apps/kubernetes/templates/cluster.yaml
index 99249c1a..1747df4e 100644
--- a/packages/apps/kubernetes/templates/cluster.yaml
+++ b/packages/apps/kubernetes/templates/cluster.yaml
@@ -1,6 +1,6 @@
 {{- $etcd := .Values._namespace.etcd }}
 {{- if not $etcd }}
-{{- fail "Kubernetes tenant requires a parent tenant with etcd enabled: set .Values.etcd=true on tenant-root (or any ancestor tenant), wait for the etcd HelmRelease to reconcile (DataStore CR appears in the tenant namespace), then retry." }}
+{{- fail "Kubernetes tenant requires a parent Tenant application with etcd: true so the etcd module deploys a DataStore CR in the tenant namespace. Set spec.etcd: true on the root Tenant (or any ancestor Tenant), wait for the etcd HelmRelease to reconcile and the DataStore CR to appear, then retry." }}
 {{- end }}
 {{- $ingress := .Values._namespace.ingress }}
 {{- $host := .Values._namespace.host }}
diff --git a/packages/apps/kubernetes/templates/csi/deploy.yaml b/packages/apps/kubernetes/templates/csi/deploy.yaml
index 2979c75c..ff170cb8 100644
--- a/packages/apps/kubernetes/templates/csi/deploy.yaml
+++ b/packages/apps/kubernetes/templates/csi/deploy.yaml
@@ -32,7 +32,12 @@ spec:
           - -c
           - |
             set -eu
+            deadline=$(( $(date +%s) + 1200 ))
             until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+              if [ "$(date +%s)" -ge "$deadline" ]; then
+                echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
+                exit 1
+              fi
               echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
               sleep 5
             done
diff --git a/packages/apps/kubernetes/templates/kccm/manager.yaml b/packages/apps/kubernetes/templates/kccm/manager.yaml
index 5c7f15a4..67321647 100644
--- a/packages/apps/kubernetes/templates/kccm/manager.yaml
+++ b/packages/apps/kubernetes/templates/kccm/manager.yaml
@@ -30,7 +30,12 @@ spec:
         - -c
         - |
           set -eu
+          deadline=$(( $(date +%s) + 1200 ))
           until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+            if [ "$(date +%s)" -ge "$deadline" ]; then
+              echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
+              exit 1
+            fi
             echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
             sleep 5
           done
diff --git a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
index 3b2b1d72..2a43e13f 100644
--- a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
+++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
@@ -116,4 +116,4 @@ tests:
         host: ""
     asserts:
       - failedTemplate:
-          errorPattern: "requires a parent tenant with etcd enabled"
+          errorPattern: "requires a parent Tenant application with etcd: true"

From b38ae605495ff30a71013037af94839c4944e815 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:05:23 +0300
Subject: [PATCH 12/82] fix(kubernetes): soft-skip cluster resources when
 tenant has no DataStore

A hard helm fail in cluster.yaml made every cold bootstrap racy: if
the parent Tenant chart had not yet populated _namespace.etcd in
cozystack-values when the Kubernetes HelmRelease first reconciled, the
fail fired, install.remediation triggered, installFailures incremented
and the new e2e remediation-guard flagged it as a bug. That directly
contradicts the race the rest of this PR is trying to close.

Replace fail with a graceful skip: render only a status-beacon
ConfigMap (test-awaiting-etcd) when etcd is empty, wrap all
CAPI/Kamaji resources in {{ if $etcd }}. The HelmRelease installs
successfully and goes Ready; flux retries on its 5m interval and picks
up the DataStore as soon as the Tenant chart finishes reconciling.

Update the helm unittest: positive test still asserts dataStoreName on
KamajiControlPlane; the negative test now asserts exactly one
ConfigMap document with status=awaiting-etcd, no Cluster / KCP /
KubevirtCluster / WorkloadMonitor rendered.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../apps/kubernetes/templates/cluster.yaml    | 37 +++++++++++++++++--
 .../tests/admin_kubeconfig_wait_test.yaml     | 19 ++++++++--
 2 files changed, 50 insertions(+), 6 deletions(-)

diff --git a/packages/apps/kubernetes/templates/cluster.yaml b/packages/apps/kubernetes/templates/cluster.yaml
index 1747df4e..12da9818 100644
--- a/packages/apps/kubernetes/templates/cluster.yaml
+++ b/packages/apps/kubernetes/templates/cluster.yaml
@@ -1,7 +1,15 @@
 {{- $etcd := .Values._namespace.etcd }}
-{{- if not $etcd }}
-{{- fail "Kubernetes tenant requires a parent Tenant application with etcd: true so the etcd module deploys a DataStore CR in the tenant namespace. Set spec.etcd: true on the root Tenant (or any ancestor Tenant), wait for the etcd HelmRelease to reconcile and the DataStore CR to appear, then retry." }}
-{{- end }}
+{{- /*
+  When $etcd is empty, the parent Tenant application has not populated
+  _namespace.etcd in cozystack-values yet - either the operator forgot to
+  set etcd: true on an ancestor Tenant, or the Tenant HelmRelease is still
+  reconciling. Either way, rendering a KamajiControlPlane with an empty
+  dataStoreName would be rejected by Kamaji's admission webhook and the
+  HelmRelease would fail to install, triggering remediation. Instead, emit
+  a single ConfigMap as a user-visible status beacon and skip the rest so
+  flux marks the HelmRelease Ready and retries its 5m reconcile loop until
+  the Tenant chart catches up.
+*/}}
 {{- $ingress := .Values._namespace.ingress }}
 {{- $host := .Values._namespace.host }}
 {{- $kubevirtmachinetemplateNames := list }}
@@ -87,6 +95,28 @@ spec:
           - name: default
             pod: {}
 {{- end }}
+{{- if not $etcd }}
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: {{ .Release.Name }}-awaiting-etcd
+  namespace: {{ .Release.Namespace }}
+  labels:
+    app.kubernetes.io/managed-by: {{ .Release.Service }}
+    app.kubernetes.io/name: kubernetes
+    app.kubernetes.io/instance: {{ .Release.Name }}
+  annotations:
+    cozystack.io/status-beacon: "true"
+data:
+  status: "awaiting-etcd"
+  message: |
+    No DataStore is available for this tenant Kubernetes cluster. The parent
+    Tenant application has not populated _namespace.etcd. Set spec.etcd: true
+    on an ancestor Tenant (usually tenant-root) and wait for its HelmRelease
+    to reconcile - this HelmRelease will pick up the DataStore on its next
+    5m reconcile loop and provision the cluster.
+{{- else }}
 ---
 apiVersion: cluster.x-k8s.io/v1beta1
 kind: Cluster
@@ -407,3 +437,4 @@ metadata:
 spec:
   {{- .spec | toYaml | nindent 2 }}
 {{- end }}
+{{- end }}
diff --git a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
index 2a43e13f..e7c03bd1 100644
--- a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
+++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
@@ -105,7 +105,12 @@ tests:
           path: spec.dataStoreName
           value: tenant-root
 
-  - it: cluster.yaml fails render when tenant has no etcd DataStore
+  - it: cluster.yaml skips Cluster resources when tenant has no etcd DataStore
+    # Must NOT fail rendering - the parent Tenant chart populates
+    # _namespace.etcd asynchronously, so rendering failures here would cause
+    # flux install remediation on every cold bootstrap. Instead, emit only a
+    # ConfigMap status beacon so the HelmRelease reports Ready while flux
+    # retries on its interval until the DataStore appears.
     template: templates/cluster.yaml
     set:
       _namespace:
@@ -115,5 +120,13 @@ tests:
         seaweedfs: ""
         host: ""
     asserts:
-      - failedTemplate:
-          errorPattern: "requires a parent Tenant application with etcd: true"
+      - hasDocuments:
+          count: 1
+      - isKind:
+          of: ConfigMap
+      - equal:
+          path: metadata.name
+          value: test-awaiting-etcd
+      - equal:
+          path: data.status
+          value: awaiting-etcd

From 1757567218ac313a030da1f82e64541a3f9bfbae Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:06:47 +0300
Subject: [PATCH 13/82] refactor(kubernetes): extract wait-for-kubeconfig init
 into shared helper

The three control-plane-side Deployments (cluster-autoscaler, kccm,
kcsi-controller) carried three copies of the same 20-line init
container. That already drifted: the CSI copy used 4-space nesting
while the other two used 2-space. Any future update to the image,
the deadline, or the poll script had to land in three places or
silently diverge.

Extract the block into a new kubernetes.waitForAdminKubeconfig helper
in templates/_helpers.tpl and include it at each call site. Tighten
the deadline from 20m to 10m so it stays strictly below the 15m
HelmRelease Install.Timeout and the CrashLoopBackOff surfaces in
dashboards before flux remediation can fire. Also clarify the wait
message so operators debugging a stuck init container do not chase
Kamaji for what is actually kubelet's optional-Secret refresh cadence.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../apps/kubernetes/templates/_helpers.tpl    | 40 +++++++++++++++++++
 .../cluster-autoscaler/deployment.yaml        | 21 +---------
 .../apps/kubernetes/templates/csi/deploy.yaml | 21 +---------
 .../kubernetes/templates/kccm/manager.yaml    | 21 +---------
 4 files changed, 43 insertions(+), 60 deletions(-)

diff --git a/packages/apps/kubernetes/templates/_helpers.tpl b/packages/apps/kubernetes/templates/_helpers.tpl
index 36c06b64..89f06934 100644
--- a/packages/apps/kubernetes/templates/_helpers.tpl
+++ b/packages/apps/kubernetes/templates/_helpers.tpl
@@ -49,3 +49,43 @@ Selector labels
 app.kubernetes.io/name: {{ include "kubernetes.name" . }}
 app.kubernetes.io/instance: {{ .Release.Name }}
 {{- end }}
+
+{{/*
+wait-for-kubeconfig init container shared by the control-plane-side
+Deployments (cluster-autoscaler, kccm, kcsi-controller) that mount the
+*-admin-kubeconfig Secret provisioned asynchronously by Kamaji. The
+Secret volume is declared optional so kubelet does not FailedMount while
+Kamaji is still bootstrapping; this container polls the mounted path and
+exits only when super-admin.svc appears, which happens after kubelet's
+optional-Secret refresh cycle.
+
+The 10m deadline stays strictly below the 15m HelmRelease
+Install.Timeout scoped to the Kubernetes Application kind so the
+CrashLoopBackOff surfaces before flux remediation fires and uninstalls
+the Cluster CR.
+
+Call site owns the surrounding volumes block; the kubeconfig volume
+must exist on the pod and mount at /etc/kubernetes/kubeconfig.
+*/}}
+{{- define "kubernetes.waitForAdminKubeconfig" -}}
+- name: wait-for-kubeconfig
+  image: "{{ .Files.Get "images/busybox.tag" | trim }}"
+  command:
+  - sh
+  - -c
+  - |
+    set -eu
+    deadline=$(( $(date +%s) + 600 ))
+    until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
+      if [ "$(date +%s)" -ge "$deadline" ]; then
+        echo "admin kubeconfig was not provisioned within 10m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
+        exit 1
+      fi
+      echo "waiting for admin kubeconfig (provisioned by Kamaji, visible after kubelet Secret refresh)..."
+      sleep 5
+    done
+  volumeMounts:
+  - name: kubeconfig
+    mountPath: /etc/kubernetes/kubeconfig
+    readOnly: true
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
index 277b6df8..348b017d 100644
--- a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
+++ b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
@@ -24,26 +24,7 @@ spec:
           operator: Exists
           effect: "NoSchedule"
       initContainers:
-      - name: wait-for-kubeconfig
-        image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
-        command:
-        - sh
-        - -c
-        - |
-          set -eu
-          deadline=$(( $(date +%s) + 1200 ))
-          until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
-            if [ "$(date +%s)" -ge "$deadline" ]; then
-              echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
-              exit 1
-            fi
-            echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
-            sleep 5
-          done
-        volumeMounts:
-        - name: kubeconfig
-          mountPath: /etc/kubernetes/kubeconfig
-          readOnly: true
+      {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }}
       containers:
       - image: "{{ $.Files.Get "images/cluster-autoscaler.tag" | trim }}"
         name: cluster-autoscaler
diff --git a/packages/apps/kubernetes/templates/csi/deploy.yaml b/packages/apps/kubernetes/templates/csi/deploy.yaml
index ff170cb8..c1af7f13 100644
--- a/packages/apps/kubernetes/templates/csi/deploy.yaml
+++ b/packages/apps/kubernetes/templates/csi/deploy.yaml
@@ -25,26 +25,7 @@ spec:
           operator: Exists
           effect: "NoSchedule"
       initContainers:
-        - name: wait-for-kubeconfig
-          image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
-          command:
-          - sh
-          - -c
-          - |
-            set -eu
-            deadline=$(( $(date +%s) + 1200 ))
-            until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
-              if [ "$(date +%s)" -ge "$deadline" ]; then
-                echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
-                exit 1
-              fi
-              echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
-              sleep 5
-            done
-          volumeMounts:
-          - name: kubeconfig
-            mountPath: /etc/kubernetes/kubeconfig
-            readOnly: true
+      {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }}
       containers:
         - name: csi-driver
           imagePullPolicy: Always
diff --git a/packages/apps/kubernetes/templates/kccm/manager.yaml b/packages/apps/kubernetes/templates/kccm/manager.yaml
index 67321647..20ac0a1e 100644
--- a/packages/apps/kubernetes/templates/kccm/manager.yaml
+++ b/packages/apps/kubernetes/templates/kccm/manager.yaml
@@ -23,26 +23,7 @@ spec:
           operator: Exists
           effect: "NoSchedule"
       initContainers:
-      - name: wait-for-kubeconfig
-        image: "{{ $.Files.Get "images/busybox.tag" | trim }}"
-        command:
-        - sh
-        - -c
-        - |
-          set -eu
-          deadline=$(( $(date +%s) + 1200 ))
-          until [ -s /etc/kubernetes/kubeconfig/super-admin.svc ]; do
-            if [ "$(date +%s)" -ge "$deadline" ]; then
-              echo "admin kubeconfig was not provisioned within 20m; exiting so the pod goes CrashLoopBackOff and surfaces in dashboards" >&2
-              exit 1
-            fi
-            echo "waiting for admin kubeconfig to be provisioned by Kamaji..."
-            sleep 5
-          done
-        volumeMounts:
-        - name: kubeconfig
-          mountPath: /etc/kubernetes/kubeconfig
-          readOnly: true
+      {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }}
       containers:
       - name: kubevirt-cloud-controller-manager
         args:

From 7b146cbe56c26511d2228c17cdb7c2201f480395 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:09:35 +0300
Subject: [PATCH 14/82] feat(api): make HelmRelease Install/Upgrade timeout
 per-Application

Replace the hardcoded r.kindName == "Kubernetes" switch in rest.go
with a config-driven path. The ApplicationDefinition CR now accepts a
release.cozystack.io/helm-install-timeout annotation that is parsed
at cozystack-api startup into config.ReleaseConfig.HelmInstallTimeout
and applied to both Install.Timeout and Upgrade.Timeout on the
rendered HelmRelease. Applications that leave the annotation unset
keep flux defaults so their failed installs remediate on the normal
cadence - only the Kubernetes kind carries the override and gets a
15m budget. New kinds with a similar race can opt in by setting the
same annotation; no rest.go patch needed.

Kubernetes-rd sets the annotation to 15m. Table-driven test in
rest_timeout_test.go covers three cases: Kubernetes with 15m, Qdrant
unset, and an arbitrary future kind with 20m - all of which pin the
Remediation.Retries == -1 contract as well.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../kubernetes-rd/cozyrds/kubernetes.yaml     |   8 +
 pkg/cmd/server/start.go                       |  35 +++--
 pkg/config/config.go                          |   8 +
 pkg/registry/apps/application/rest.go         |  21 +--
 .../apps/application/rest_timeout_test.go     | 143 ++++++++++--------
 5 files changed, 133 insertions(+), 82 deletions(-)

diff --git a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
index 5e9e8f94..1128abe8 100644
--- a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
+++ b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
@@ -2,6 +2,14 @@ apiVersion: cozystack.io/v1alpha1
 kind: ApplicationDefinition
 metadata:
   name: kubernetes
+  annotations:
+    # Kubernetes tenants boot a Kamaji control plane whose admin-kubeconfig
+    # Secret is provisioned asynchronously. Cold Kamaji start (image pull +
+    # etcd + apiserver Ready) plus admin-kubeconfig generation can exceed
+    # Flux helm-controller's default wait budget, causing remediation loops
+    # that uninstall the Cluster CR. This override applied by cozystack-api
+    # to the HelmRelease Spec.Install.Timeout and Spec.Upgrade.Timeout.
+    release.cozystack.io/helm-install-timeout: "15m"
 spec:
   application:
     kind: Kubernetes
diff --git a/pkg/cmd/server/start.go b/pkg/cmd/server/start.go
index f7791332..3e905735 100644
--- a/pkg/cmd/server/start.go
+++ b/pkg/cmd/server/start.go
@@ -160,6 +160,31 @@ func (o *CozyServerOptions) Complete() error {
 	// Convert to ResourceConfig
 	o.ResourceConfig = &config.ResourceConfig{}
 	for _, crd := range crdList.Items {
+		release := config.ReleaseConfig{
+			Prefix: crd.Spec.Release.Prefix,
+			Labels: crd.Spec.Release.Labels,
+			ChartRef: config.ChartRefConfig{
+				Kind:      crd.Spec.Release.ChartRef.Kind,
+				Name:      crd.Spec.Release.ChartRef.Name,
+				Namespace: crd.Spec.Release.ChartRef.Namespace,
+			},
+		}
+		// Per-Application HelmRelease Install/Upgrade timeout. Applications
+		// whose parent chart contains asynchronously-provisioned resources
+		// the chart itself depends on (for example, the Kamaji-provisioned
+		// admin-kubeconfig Secret for Kubernetes tenants) need a longer
+		// wait budget than the Flux default. Consumed by the REST storage
+		// layer when building the HelmRelease Spec.
+		if raw, ok := crd.Annotations["release.cozystack.io/helm-install-timeout"]; ok && raw != "" {
+			d, err := time.ParseDuration(raw)
+			if err != nil {
+				return fmt.Errorf(
+					"ApplicationDefinition %q has invalid release.cozystack.io/helm-install-timeout %q: %w",
+					crd.Name, raw, err,
+				)
+			}
+			release.HelmInstallTimeout = d
+		}
 		resource := config.Resource{
 			Application: config.ApplicationConfig{
 				Kind:          crd.Spec.Application.Kind,
@@ -168,15 +193,7 @@ func (o *CozyServerOptions) Complete() error {
 				ShortNames:    []string{}, // TODO: implement shortnames
 				OpenAPISchema: crd.Spec.Application.OpenAPISchema,
 			},
-			Release: config.ReleaseConfig{
-				Prefix: crd.Spec.Release.Prefix,
-				Labels: crd.Spec.Release.Labels,
-				ChartRef: config.ChartRefConfig{
-					Kind:      crd.Spec.Release.ChartRef.Kind,
-					Name:      crd.Spec.Release.ChartRef.Name,
-					Namespace: crd.Spec.Release.ChartRef.Namespace,
-				},
-			},
+			Release: release,
 		}
 		o.ResourceConfig.Resources = append(o.ResourceConfig.Resources, resource)
 	}
diff --git a/pkg/config/config.go b/pkg/config/config.go
index 1e123e2c..16cf4f0c 100644
--- a/pkg/config/config.go
+++ b/pkg/config/config.go
@@ -16,6 +16,8 @@ limitations under the License.
 
 package config
 
+import "time"
+
 // ResourceConfig represents the structure of the configuration file.
 type ResourceConfig struct {
 	Resources []Resource `yaml:"resources"`
@@ -41,6 +43,12 @@ type ReleaseConfig struct {
 	Prefix   string            `yaml:"prefix"`
 	Labels   map[string]string `yaml:"labels"`
 	ChartRef ChartRefConfig    `yaml:"chartRef"`
+	// HelmInstallTimeout overrides the Flux HelmRelease Install.Timeout and
+	// Upgrade.Timeout for this Application kind. When zero, flux defaults
+	// apply. Populated from the
+	// release.cozystack.io/helm-install-timeout annotation on the
+	// ApplicationDefinition at start-up.
+	HelmInstallTimeout time.Duration `yaml:"helmInstallTimeout,omitempty"`
 }
 
 // ChartRefConfig references a Flux source artifact for the Helm chart.
diff --git a/pkg/registry/apps/application/rest.go b/pkg/registry/apps/application/rest.go
index 9b92d2c0..406d3738 100644
--- a/pkg/registry/apps/application/rest.go
+++ b/pkg/registry/apps/application/rest.go
@@ -1528,16 +1528,17 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (*
 		},
 	}
 
-	// The Kubernetes Application's parent chart creates CAPI/Kamaji resources
-	// whose admin-kubeconfig Secret is provisioned asynchronously and mounted
-	// by Deployments in the same chart. On a cold node, Kamaji control-plane
-	// bootstrap routinely exceeds flux helm-controller's default wait budget,
-	// so install remediation loops uninstall the Cluster CR and churn. Extend
-	// the wait budget to 15m for Kubernetes only - other Application kinds
-	// without this race keep flux defaults, so their failed installs do not
-	// linger unnecessarily before remediation fires.
-	if r.kindName == "Kubernetes" {
-		timeout := metav1.Duration{Duration: 15 * time.Minute}
+	// Per-Application HelmRelease wait budget. When an ApplicationDefinition
+	// sets release.cozystack.io/helm-install-timeout, the annotation is
+	// parsed at startup into ReleaseConfig.HelmInstallTimeout and applied
+	// to both Install and Upgrade here. Applications that leave it unset
+	// (the common case) keep flux defaults, so their failed installs
+	// remediate on the normal cadence. Needed for the Kubernetes kind
+	// because its parent chart contains CAPI/Kamaji resources whose
+	// admin-kubeconfig Secret is provisioned asynchronously and Kamaji
+	// cold-start routinely exceeds flux's default wait budget.
+	if r.releaseConfig.HelmInstallTimeout > 0 {
+		timeout := metav1.Duration{Duration: r.releaseConfig.HelmInstallTimeout}
 		helmRelease.Spec.Install.Timeout = &timeout
 		helmRelease.Spec.Upgrade.Timeout = &timeout
 	}
diff --git a/pkg/registry/apps/application/rest_timeout_test.go b/pkg/registry/apps/application/rest_timeout_test.go
index fa6d5427..8474d1cb 100644
--- a/pkg/registry/apps/application/rest_timeout_test.go
+++ b/pkg/registry/apps/application/rest_timeout_test.go
@@ -10,7 +10,7 @@ import (
 	"github.com/cozystack/cozystack/pkg/config"
 )
 
-func newRESTForKind(kind, prefix string) *REST {
+func newRESTForTimeout(kind, prefix string, helmInstallTimeout time.Duration) *REST {
 	return &REST{
 		kindName: kind,
 		releaseConfig: config.ReleaseConfig{
@@ -20,76 +20,93 @@ func newRESTForKind(kind, prefix string) *REST {
 				Name:      "x",
 				Namespace: "cozy-system",
 			},
+			HelmInstallTimeout: helmInstallTimeout,
 		},
 	}
 }
 
-func TestConvertApplicationToHelmRelease_KubernetesKindGetsLongTimeout(t *testing.T) {
-	r := newRESTForKind("Kubernetes", "kubernetes-")
-	app := &appsv1alpha1.Application{
-		ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
+// Table-driven: every Application kind carries a per-CRD HelmRelease wait
+// budget. The Kubernetes kind's parent chart contains CAPI/Kamaji resources
+// whose admin-kubeconfig Secret is provisioned asynchronously, so its
+// ApplicationDefinition sets release.cozystack.io/helm-install-timeout=15m
+// (or longer). Other kinds leave the annotation unset and keep flux defaults
+// so their failed installs remediate on the normal cadence. The test must
+// cover both paths: a kind with the timeout set and one without.
+func TestConvertApplicationToHelmRelease_AppliesReleaseConfigTimeout(t *testing.T) {
+	cases := []struct {
+		name       string
+		kind       string
+		prefix     string
+		configured time.Duration
+		wantSet    bool
+	}{
+		{
+			name:       "Kubernetes kind with 15m configured gets Install and Upgrade Timeout",
+			kind:       "Kubernetes",
+			prefix:     "kubernetes-",
+			configured: 15 * time.Minute,
+			wantSet:    true,
+		},
+		{
+			name:       "Qdrant kind without configured timeout keeps flux defaults",
+			kind:       "Qdrant",
+			prefix:     "qdrant-",
+			configured: 0,
+			wantSet:    false,
+		},
+		{
+			name:       "arbitrary future kind with 20m configured gets 20m",
+			kind:       "TalosCluster",
+			prefix:     "talos-",
+			configured: 20 * time.Minute,
+			wantSet:    true,
+		},
 	}
 
-	hr, err := r.convertApplicationToHelmRelease(app)
-	if err != nil {
-		t.Fatalf("convertApplicationToHelmRelease returned error: %v", err)
-	}
+	for _, tc := range cases {
+		t.Run(tc.name, func(t *testing.T) {
+			r := newRESTForTimeout(tc.kind, tc.prefix, tc.configured)
+			app := &appsv1alpha1.Application{
+				ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
+			}
 
-	if hr.Spec.Install == nil {
-		t.Fatal("Spec.Install must not be nil")
-	}
-	if hr.Spec.Install.Timeout == nil {
-		t.Fatal("Spec.Install.Timeout must be set for Kubernetes kind")
-	}
-	if hr.Spec.Install.Timeout.Duration < 15*time.Minute {
-		t.Errorf("Spec.Install.Timeout must be >= 15m for Kubernetes, got %v", hr.Spec.Install.Timeout.Duration)
-	}
+			hr, err := r.convertApplicationToHelmRelease(app)
+			if err != nil {
+				t.Fatalf("convertApplicationToHelmRelease returned error: %v", err)
+			}
 
-	if hr.Spec.Upgrade == nil {
-		t.Fatal("Spec.Upgrade must not be nil")
-	}
-	if hr.Spec.Upgrade.Timeout == nil {
-		t.Fatal("Spec.Upgrade.Timeout must be set for Kubernetes kind")
-	}
-	if hr.Spec.Upgrade.Timeout.Duration < 15*time.Minute {
-		t.Errorf("Spec.Upgrade.Timeout must be >= 15m for Kubernetes, got %v", hr.Spec.Upgrade.Timeout.Duration)
-	}
+			if hr.Spec.Install == nil || hr.Spec.Upgrade == nil {
+				t.Fatalf("Spec.Install/Upgrade must be non-nil")
+			}
 
-	if hr.Spec.Install.Remediation == nil || hr.Spec.Install.Remediation.Retries != -1 {
-		t.Errorf("Spec.Install.Remediation.Retries must remain -1, got %+v", hr.Spec.Install.Remediation)
-	}
-	if hr.Spec.Upgrade.Remediation == nil || hr.Spec.Upgrade.Remediation.Retries != -1 {
-		t.Errorf("Spec.Upgrade.Remediation.Retries must remain -1, got %+v", hr.Spec.Upgrade.Remediation)
-	}
-}
-
-func TestConvertApplicationToHelmRelease_NonKubernetesKindKeepsFluxDefaults(t *testing.T) {
-	// For Applications whose parent chart has no admin-kubeconfig race
-	// (Qdrant, MongoDB, Postgres, etc.), do NOT extend the helm-wait
-	// budget - otherwise failed installs would block three times as long
-	// before Flux starts remediating.
-	r := newRESTForKind("Qdrant", "qdrant-")
-	app := &appsv1alpha1.Application{
-		ObjectMeta: metav1.ObjectMeta{Name: "example", Namespace: "tenant-root"},
-	}
-
-	hr, err := r.convertApplicationToHelmRelease(app)
-	if err != nil {
-		t.Fatalf("convertApplicationToHelmRelease returned error: %v", err)
-	}
-
-	if hr.Spec.Install != nil && hr.Spec.Install.Timeout != nil {
-		t.Errorf("Spec.Install.Timeout must be unset for non-Kubernetes kinds, got %v", hr.Spec.Install.Timeout.Duration)
-	}
-	if hr.Spec.Upgrade != nil && hr.Spec.Upgrade.Timeout != nil {
-		t.Errorf("Spec.Upgrade.Timeout must be unset for non-Kubernetes kinds, got %v", hr.Spec.Upgrade.Timeout.Duration)
-	}
-
-	// But remediation must still be -1 across the board.
-	if hr.Spec.Install.Remediation == nil || hr.Spec.Install.Remediation.Retries != -1 {
-		t.Errorf("Spec.Install.Remediation.Retries must remain -1, got %+v", hr.Spec.Install.Remediation)
-	}
-	if hr.Spec.Upgrade.Remediation == nil || hr.Spec.Upgrade.Remediation.Retries != -1 {
-		t.Errorf("Spec.Upgrade.Remediation.Retries must remain -1, got %+v", hr.Spec.Upgrade.Remediation)
+			if tc.wantSet {
+				if hr.Spec.Install.Timeout == nil {
+					t.Fatalf("Spec.Install.Timeout must be set when HelmInstallTimeout=%v", tc.configured)
+				}
+				if hr.Spec.Install.Timeout.Duration != tc.configured {
+					t.Errorf("Spec.Install.Timeout = %v, want %v", hr.Spec.Install.Timeout.Duration, tc.configured)
+				}
+				if hr.Spec.Upgrade.Timeout == nil {
+					t.Fatalf("Spec.Upgrade.Timeout must be set when HelmInstallTimeout=%v", tc.configured)
+				}
+				if hr.Spec.Upgrade.Timeout.Duration != tc.configured {
+					t.Errorf("Spec.Upgrade.Timeout = %v, want %v", hr.Spec.Upgrade.Timeout.Duration, tc.configured)
+				}
+			} else {
+				if hr.Spec.Install.Timeout != nil {
+					t.Errorf("Spec.Install.Timeout must be nil when HelmInstallTimeout=0, got %v", hr.Spec.Install.Timeout.Duration)
+				}
+				if hr.Spec.Upgrade.Timeout != nil {
+					t.Errorf("Spec.Upgrade.Timeout must be nil when HelmInstallTimeout=0, got %v", hr.Spec.Upgrade.Timeout.Duration)
+				}
+			}
+
+			if hr.Spec.Install.Remediation == nil || hr.Spec.Install.Remediation.Retries != -1 {
+				t.Errorf("Spec.Install.Remediation.Retries must remain -1, got %+v", hr.Spec.Install.Remediation)
+			}
+			if hr.Spec.Upgrade.Remediation == nil || hr.Spec.Upgrade.Remediation.Retries != -1 {
+				t.Errorf("Spec.Upgrade.Remediation.Retries must remain -1, got %+v", hr.Spec.Upgrade.Remediation)
+			}
+		})
 	}
 }

From bc5473d4fc9bb1b43be79ea5eb8d370d9e6c00b7 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:11:38 +0300
Subject: [PATCH 15/82] test(kubernetes): chart-wide invariant for
 admin-kubeconfig guards

The per-template unittests in packages/apps/kubernetes/tests/ assert
that cluster-autoscaler, kccm, and the csi controller each mount the
admin-kubeconfig Secret optional and carry the wait-for-kubeconfig
init. That locks in today's three Deployments by name - a fourth
Deployment that mounts the same Secret but forgets the guard would
slip past them.

Add a bats-unit test that renders the entire chart, enumerates every
Deployment whose spec mounts a Secret ending in -admin-kubeconfig, and
asserts optional:true plus wait-for-kubeconfig init on all of them.
Verified by temporarily removing optional:true from csi/deploy.yaml:
the test correctly flagged invariant-kcsi-controller as an offender.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/admin-kubeconfig-invariant.bats | 72 ++++++++++++++++++++++++++++
 1 file changed, 72 insertions(+)
 create mode 100644 hack/admin-kubeconfig-invariant.bats

diff --git a/hack/admin-kubeconfig-invariant.bats b/hack/admin-kubeconfig-invariant.bats
new file mode 100644
index 00000000..d699de83
--- /dev/null
+++ b/hack/admin-kubeconfig-invariant.bats
@@ -0,0 +1,72 @@
+#!/usr/bin/env bats
+# -----------------------------------------------------------------------------
+# Chart-wide invariant for packages/apps/kubernetes:
+#
+# Every Deployment in this chart that mounts <release>-admin-kubeconfig as a
+# Secret volume MUST:
+#   - declare that volume optional: true (so kubelet does not FailedMount
+#     while Kamaji is still provisioning the Secret), AND
+#   - include the wait-for-kubeconfig init container (so the pod becomes
+#     Ready only after Kamaji publishes the Secret).
+#
+# The per-template unittests in packages/apps/kubernetes/tests/ lock in
+# today's three Deployments (cluster-autoscaler, kccm, csi controller) by
+# name. This invariant is stricter: any future Deployment added to this
+# chart that mounts the same Secret but forgets the guard will fail here.
+#
+# Requires: helm, yq (mikefarah v4+), jq. All three are available on the
+# project's CI runners and on the maintainer workstation.
+# -----------------------------------------------------------------------------
+
+@test "every Deployment mounting admin-kubeconfig has optional:true and wait-for-kubeconfig init" {
+  values_file="packages/apps/kubernetes/tests/values-ci.yaml"
+  [ -f "$values_file" ]
+
+  tmp=$(mktemp -d)
+  trap 'rm -rf "$tmp"' EXIT
+
+  helm template invariant packages/apps/kubernetes \
+    --namespace tenant-root \
+    --values "$values_file" \
+    2>/dev/null > "$tmp/rendered.yaml"
+
+  # yq streams one JSON object per input document. jq -s slurps the stream
+  # into an array so we can treat all Deployments as a single collection.
+  yq --output-format=json eval-all '.' "$tmp/rendered.yaml" \
+    | jq -s --raw-output '
+        map(select(.kind == "Deployment")) |
+        map({
+          name: .metadata.name,
+          volumes: (.spec.template.spec.volumes // []),
+          initNames: ((.spec.template.spec.initContainers // []) | map(.name)),
+        }) |
+        map(
+          .name as $n |
+          .initNames as $ins |
+          (.volumes[] | select(.secret.secretName | test("-admin-kubeconfig$")?))
+          | {
+              name: $n,
+              optional: (.secret.optional == true),
+              hasInit: ($ins | index("wait-for-kubeconfig") != null),
+            }
+        )
+      ' > "$tmp/summary.json"
+
+  # At least one Deployment must match; if a refactor removes every
+  # admin-kubeconfig volume from this chart, the test must be updated
+  # deliberately rather than silently passing.
+  matched=$(jq 'length' "$tmp/summary.json")
+  [ "$matched" -ge 1 ]
+
+  offenders=$(jq --raw-output '.[] | select(.optional != true or .hasInit != true) | .name' "$tmp/summary.json")
+
+  if [ -n "$offenders" ]; then
+    echo "Deployments mounting *-admin-kubeconfig without optional:true + wait-for-kubeconfig init:" >&2
+    echo "$offenders" >&2
+    echo "Full summary:" >&2
+    cat "$tmp/summary.json" >&2
+    exit 1
+  fi
+
+  echo "Invariant holds for $matched Deployment(s)"
+}

From 03426fbd718029e40d6680f299189e97f915a614 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:13:06 +0300
Subject: [PATCH 16/82] test(hack): pin HelmRelease v2 status shape used by
 remediation guard

run-kubernetes.sh extracts .status.installFailures and
.status.upgradeFailures via kubectl -o jsonpath. If a future flux
release renames the counters, kubectl returns empty, the guard reports
no cycle, and e2e silently misses real remediation loops.

Add a bats unit test that feeds a pinned HelmRelease v2 status snippet
through the same jsonpath and asserts the extraction still yields the
expected values. Also leave a pointer comment in run-kubernetes.sh so a
future flux bump surfaces the version coupling in review.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/e2e-apps/run-kubernetes.sh |  7 ++++++
 hack/remediation-guard.bats     | 44 +++++++++++++++++++++++++++++++++
 2 files changed, 51 insertions(+)

diff --git a/hack/e2e-apps/run-kubernetes.sh b/hack/e2e-apps/run-kubernetes.sh
index 811f42a7..e6002b53 100644
--- a/hack/e2e-apps/run-kubernetes.sh
+++ b/hack/e2e-apps/run-kubernetes.sh
@@ -326,6 +326,13 @@ EOF
   # A non-zero installFailures/upgradeFailures indicates the helm-wait budget expired while
   # admin-kubeconfig was still being provisioned, which would trigger uninstall remediation
   # and churn the Cluster CR.
+  # Flux helm-controller v2 status shape: .status.installFailures and
+  # .status.upgradeFailures are counters populated by the controller on
+  # every failed install/upgrade. If a future flux release renames them,
+  # kubectl returns the empty string and the guard silently passes. The
+  # shape is pinned by hack/remediation-guard.bats (see that file for
+  # details), and the vendored API types live under
+  # vendor/github.com/fluxcd/helm-controller/api/v2.
   install_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.installFailures}')
   upgrade_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.upgradeFailures}')
   if helmrelease_has_remediation_cycle "${install_failures}" "${upgrade_failures}"; then
diff --git a/hack/remediation-guard.bats b/hack/remediation-guard.bats
index 64aa83af..9c33452a 100644
--- a/hack/remediation-guard.bats
+++ b/hack/remediation-guard.bats
@@ -85,3 +85,47 @@
     helmrelease_has_remediation_cycle "3" "5" || rc=$?
     [ "$rc" -eq 0 ]
 }
+
+@test "installFailures and upgradeFailures extraction pins HR v2 status shape" {
+    # Pins the Flux HelmRelease v2 status shape that run-kubernetes.sh relies
+    # on. If a future Flux version renames .status.installFailures (or
+    # .status.upgradeFailures), kubectl get -o jsonpath returns an empty
+    # string, the guard quietly says "no cycle", and real remediation loops
+    # slip past the e2e assertion.
+    #
+    # This test uses yq to read the exact path used in the e2e script. yq
+    # evaluates the same json-ish jsonpath against a pinned HR snippet, so
+    # the test fails loudly if the field ever disappears or moves. Cross
+    # reference: vendor/github.com/fluxcd/helm-controller/api/v2/ status
+    # struct field tags.
+    tmp=$(mktemp -d)
+    trap 'rm -rf "$tmp"' EXIT
+
+    cat > "$tmp/hr.yaml" <<'YAML'
+apiVersion: helm.toolkit.fluxcd.io/v2
+kind: HelmRelease
+metadata:
+  name: kubernetes-test
+  namespace: tenant-test
+spec:
+  interval: 5m
+status:
+  installFailures: 2
+  upgradeFailures: 0
+  conditions:
+    - type: Ready
+      status: "False"
+      reason: UninstallSucceeded
+YAML
+
+    install_failures=$(yq '.status.installFailures' "$tmp/hr.yaml")
+    upgrade_failures=$(yq '.status.upgradeFailures' "$tmp/hr.yaml")
+
+    [ "$install_failures" = "2" ]
+    [ "$upgrade_failures" = "0" ]
+
+    . hack/e2e-apps/remediation-guard.sh
+    rc=0
+    helmrelease_has_remediation_cycle "$install_failures" "$upgrade_failures" || rc=$?
+    [ "$rc" -eq 0 ]
+}

From 12632c60c76b77639c8e2edd7ab5bdb248398a7e Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:25:50 +0300
Subject: [PATCH 17/82] fix(kubernetes): gate CP-side Deployments on tenant
 etcd DataStore

The soft-skip wrap in cluster.yaml only silenced Cluster and
KamajiControlPlane rendering when _namespace.etcd is empty. The three
CP-side Deployments (cluster-autoscaler, kccm, kcsi-controller) still
rendered, their wait-for-kubeconfig init containers CrashLoopBackOff'd
forever (no KamajiControlPlane = no admin-kubeconfig Secret),
HelmRelease hit its 15m wait timeout and triggered the very install
remediation cycle the rest of this PR prevents. Self-contradiction.

Wrap each of the three Deployment templates in
{{- if .Values._namespace.etcd }}...{{- end }} so they render only
when there is a DataStore to back them. Add an invariant bats test
that renders the whole chart with etcd empty and asserts zero
Deployments reference *-admin-kubeconfig.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/admin-kubeconfig-invariant.bats          | 45 +++++++++++++++++++
 .../cluster-autoscaler/deployment.yaml        | 12 +++++
 .../apps/kubernetes/templates/csi/deploy.yaml |  2 +
 .../kubernetes/templates/kccm/manager.yaml    |  2 +
 4 files changed, 61 insertions(+)

diff --git a/hack/admin-kubeconfig-invariant.bats b/hack/admin-kubeconfig-invariant.bats
index d699de83..c027e2c4 100644
--- a/hack/admin-kubeconfig-invariant.bats
+++ b/hack/admin-kubeconfig-invariant.bats
@@ -70,3 +70,48 @@
 
   echo "Invariant holds for $matched Deployment(s)"
 }
+
+@test "chart emits zero admin-kubeconfig Deployments when tenant has no etcd DataStore" {
+  # Without a DataStore (parent Tenant has not populated _namespace.etcd yet)
+  # the control-plane-side Deployments must NOT render at all. If they did,
+  # the wait-for-kubeconfig init would CrashLoopBackOff indefinitely - there
+  # would be no KamajiControlPlane to provision the Secret - consuming the
+  # HelmRelease wait budget and triggering exactly the remediation cycle the
+  # rest of this chart tries to avoid. This test renders the whole chart
+  # with etcd empty and asserts no Deployment references the admin-kubeconfig
+  # Secret.
+  tmp=$(mktemp -d)
+  trap 'rm -rf "$tmp"' EXIT
+
+  helm template invariant packages/apps/kubernetes \
+    --namespace tenant-root \
+    --set _namespace.etcd="" \
+    --set _namespace.monitoring="" \
+    --set _namespace.ingress="" \
+    --set _namespace.seaweedfs="" \
+    --set _namespace.host="" \
+    --set _cluster.cluster-domain=cozy.local \
+    --set 'nodeGroups=null' \
+    2>/dev/null > "$tmp/rendered.yaml"
+
+  matched=$(
+    yq --output-format=json eval-all '.' "$tmp/rendered.yaml" \
+      | jq -s '
+          map(select(.kind == "Deployment")) |
+          map(select(
+            (.spec.template.spec.volumes // [])
+              | any(.secret.secretName | test("-admin-kubeconfig$")?)
+          )) |
+          length
+        '
+  )
+
+  if [ "$matched" -ne 0 ]; then
+    echo "Expected zero Deployments mounting *-admin-kubeconfig when etcd is empty, got $matched:" >&2
+    yq --output-format=json eval-all '.' "$tmp/rendered.yaml" \
+      | jq -s 'map(select(.kind == "Deployment") | .metadata.name)' >&2
+    exit 1
+  fi
+
+  echo "No admin-kubeconfig Deployments rendered for empty etcd (as expected)"
+}
diff --git a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
index 348b017d..298d86db 100644
--- a/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
+++ b/packages/apps/kubernetes/templates/cluster-autoscaler/deployment.yaml
@@ -1,3 +1,14 @@
+{{- /*
+  Gate the control-plane-side workloads on the parent tenant having an etcd
+  DataStore. Without it no KamajiControlPlane is ever created, Kamaji never
+  provisions -admin-kubeconfig, and rendering these Deployments would cause
+  the wait-for-kubeconfig init to CrashLoopBackOff indefinitely, consuming
+  the parent HelmRelease install timeout and triggering the very uninstall
+  remediation cycle this chart is supposed to avoid. Rendering them only
+  when $etcd is set keeps the HelmRelease Ready while flux retries on its
+  interval and picks up the DataStore as soon as the Tenant chart finishes.
+*/}}
+{{- if .Values._namespace.etcd }}
 ---
 apiVersion: apps/v1
 kind: Deployment
@@ -108,3 +119,4 @@ rules:
     - list
     - update
     - watch
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/csi/deploy.yaml b/packages/apps/kubernetes/templates/csi/deploy.yaml
index c1af7f13..de62104c 100644
--- a/packages/apps/kubernetes/templates/csi/deploy.yaml
+++ b/packages/apps/kubernetes/templates/csi/deploy.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 kind: Deployment
 apiVersion: apps/v1
 metadata:
@@ -238,3 +239,4 @@ spec:
             secretName: {{ .Release.Name }}-admin-kubeconfig
             optional: true
           name: kubeconfig
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/kccm/manager.yaml b/packages/apps/kubernetes/templates/kccm/manager.yaml
index 20ac0a1e..bd9e2798 100644
--- a/packages/apps/kubernetes/templates/kccm/manager.yaml
+++ b/packages/apps/kubernetes/templates/kccm/manager.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 apiVersion: apps/v1
 kind: Deployment
 metadata:
@@ -60,3 +61,4 @@ spec:
           optional: true
         name: kubeconfig
       serviceAccountName: {{ .Release.Name }}-kccm
+{{- end }}

From 1ddd2aaea97024f6803c60d29a946273b55e4185 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:28:11 +0300
Subject: [PATCH 18/82] fix(hack): detect remediation via status.history, not
 transient counters

Flux helm-controller's ClearFailures() zeroes installFailures and
upgradeFailures on every successful reconciliation (see the upstream
HelmReleaseStatus method). The previous guard ran after the HelmRelease
was Ready, at which point the counters were always 0 - the assertion
was vacuous and would have passed against a reverted fix.

Switch to .status.history, which retains per-revision release
Snapshots that survive a subsequent successful reconciliation. A
remediation cycle leaves behind a Snapshot with status=uninstalled
(the install-remediation code path) or status=failed (Helm release
failure that remediation then uninstalled). Either one signals the
race actually fired.

Rewrite the bats unit tests to cover: empty history, deployed-only,
deployed+superseded (happy path - not detected), single failed,
single uninstalled, uninstalled-then-deployed, and deployed-then-failed
(all detected). The pinned-shape test feeds a realistic HR status
snippet through yq the same way run-kubernetes.sh does via kubectl
-o jsonpath.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/e2e-apps/remediation-guard.sh |  45 +++++---
 hack/e2e-apps/run-kubernetes.sh    |  24 +++--
 hack/remediation-guard.bats        | 158 ++++++++++++++---------------
 3 files changed, 120 insertions(+), 107 deletions(-)

diff --git a/hack/e2e-apps/remediation-guard.sh b/hack/e2e-apps/remediation-guard.sh
index 6da69f13..38463310 100644
--- a/hack/e2e-apps/remediation-guard.sh
+++ b/hack/e2e-apps/remediation-guard.sh
@@ -1,24 +1,39 @@
 # Helpers for asserting that a Flux HelmRelease did not fall into an
 # install/upgrade remediation cycle during an e2e run.
 #
-# A non-zero installFailures/upgradeFailures counter means flux
-# helm-controller hit its wait timeout, ran remediation (uninstall),
-# and re-installed. That is exactly the race this guard is meant to
-# catch, so the function returns success (0) when a cycle is detected
-# and failure (1) otherwise.
+# Background: Flux helm-controller's ClearFailures() zeroes
+# .status.installFailures / .status.upgradeFailures on every successful
+# reconciliation (see the upstream ClearFailures method on
+# HelmReleaseStatus). That makes those counters useless for a guard that
+# runs after the HelmRelease has reached Ready - the values are always 0.
 #
-# Both arguments may be empty strings, the literal "0", or a positive
-# integer. Shell's && and || have equal precedence with left-to-right
-# associativity, so each half of the disjunction is grouped explicitly
-# to avoid (A && B) || C && D parsing that masks the common
-# install_failures=1, upgrade_failures="" case.
+# What survives a successful reconciliation is .status.history, a bounded
+# list of release Snapshots. Each Snapshot carries a status field that
+# tracks the Helm release state: deployed, superseded, failed, uninstalled,
+# and so on. A remediation cycle leaves the footprint behind: a snapshot
+# with status "uninstalled" (from install/upgrade remediation) or "failed"
+# (Helm release failure that remediation then uninstalled). Those stay in
+# history even after a subsequent successful reinstall.
+#
+# helmrelease_has_remediation_cycle takes a newline-delimited list of
+# snapshot statuses (whatever the caller extracted via kubectl -o jsonpath
+# or equivalent) and returns 0 (detected) when any entry is "failed" or
+# "uninstalled", 1 otherwise. Empty input is treated as "no history yet,
+# no cycle observed".
 
 helmrelease_has_remediation_cycle() {
-    install_failures="$1"
-    upgrade_failures="$2"
-    if { [ -n "${install_failures}" ] && [ "${install_failures}" != "0" ]; } || \
-       { [ -n "${upgrade_failures}" ] && [ "${upgrade_failures}" != "0" ]; }; then
-        return 0
+    statuses="$1"
+    if [ -z "${statuses}" ]; then
+        return 1
     fi
+    while IFS= read -r status; do
+        case "${status}" in
+            failed|uninstalled)
+                return 0
+                ;;
+        esac
+    done <<EOF
+${statuses}
+EOF
     return 1
 }
diff --git a/hack/e2e-apps/run-kubernetes.sh b/hack/e2e-apps/run-kubernetes.sh
index e6002b53..64b1ae90 100644
--- a/hack/e2e-apps/run-kubernetes.sh
+++ b/hack/e2e-apps/run-kubernetes.sh
@@ -326,17 +326,19 @@ EOF
   # A non-zero installFailures/upgradeFailures indicates the helm-wait budget expired while
   # admin-kubeconfig was still being provisioned, which would trigger uninstall remediation
   # and churn the Cluster CR.
-  # Flux helm-controller v2 status shape: .status.installFailures and
-  # .status.upgradeFailures are counters populated by the controller on
-  # every failed install/upgrade. If a future flux release renames them,
-  # kubectl returns the empty string and the guard silently passes. The
-  # shape is pinned by hack/remediation-guard.bats (see that file for
-  # details), and the vendored API types live under
-  # vendor/github.com/fluxcd/helm-controller/api/v2.
-  install_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.installFailures}')
-  upgrade_failures=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" -ojsonpath='{.status.upgradeFailures}')
-  if helmrelease_has_remediation_cycle "${install_failures}" "${upgrade_failures}"; then
-    echo "Parent HelmRelease entered remediation cycle: installFailures=${install_failures:-0}, upgradeFailures=${upgrade_failures:-0}" >&2
+  # Flux helm-controller v2 retains per-revision release Snapshots in
+  # .status.history; each Snapshot's .status reflects the Helm release
+  # state (deployed/superseded/failed/uninstalled). A remediation cycle
+  # leaves a "failed" or "uninstalled" entry behind that survives a later
+  # successful reinstall, unlike the installFailures/upgradeFailures
+  # counters (which ClearFailures zeroes on every successful reconcile).
+  # The shape is pinned by hack/remediation-guard.bats; the upstream
+  # types are github.com/fluxcd/helm-controller/api v2 Snapshot.
+  history_statuses=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" \
+    -ojsonpath='{range .status.history[*]}{.status}{"\n"}{end}')
+  if helmrelease_has_remediation_cycle "${history_statuses}"; then
+    echo "Parent HelmRelease entered remediation cycle. History statuses:" >&2
+    printf '%s\n' "${history_statuses}" >&2
     kubectl -n tenant-test describe hr "kubernetes-${test_name}" >&2
     exit 1
   fi
diff --git a/hack/remediation-guard.bats b/hack/remediation-guard.bats
index 9c33452a..30e4e04c 100644
--- a/hack/remediation-guard.bats
+++ b/hack/remediation-guard.bats
@@ -2,17 +2,15 @@
 # -----------------------------------------------------------------------------
 # Unit tests for hack/e2e-apps/remediation-guard.sh
 #
-# helmrelease_has_remediation_cycle is consumed from e2e tests to assert that
-# the parent HelmRelease did not hit flux helm-controller's wait timeout and
-# enter uninstall remediation. The function accepts two arguments (values of
-# .status.installFailures and .status.upgradeFailures) and returns 0 when a
-# remediation cycle is detected, 1 otherwise.
+# helmrelease_has_remediation_cycle takes a newline-delimited list of
+# HelmRelease history snapshot status values (deployed/superseded/failed/
+# uninstalled/...) and returns 0 when any entry is "failed" or "uninstalled"
+# (meaning flux helm-controller performed install/upgrade remediation).
 #
-# Each argument can be empty (controller never populated the field), "0"
-# (populated but never failed), or a positive integer. Shell's && and ||
-# have equal precedence with left-to-right associativity, which used to
-# break this check on the most common failure mode - install_failures=1
-# and upgrade_failures="". These tests pin the correct behavior.
+# This is used by the e2e script after the HelmRelease reaches Ready. The
+# failure/upgrade counters (.status.installFailures / .status.upgradeFailures)
+# are useless there because flux's ClearFailures zeroes them on successful
+# reconciliation; .status.history retains the snapshot trail.
 #
 # cozytest.sh's awk parser recognizes only @test blocks and a bare `}` on
 # its own line; there is no bats `run` or `$status`. Assertions are
@@ -21,83 +19,76 @@
 # Run with: hack/cozytest.sh hack/remediation-guard.bats
 # -----------------------------------------------------------------------------
 
-@test "no counters set returns not-detected" {
+@test "empty history returns not-detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "" "" || rc=$?
-    [ "$rc" -eq 1 ]
+    if helmrelease_has_remediation_cycle ""; then
+        echo "expected not-detected for empty history" >&2
+        exit 1
+    fi
 }
 
-@test "both counters zero returns not-detected" {
+@test "single deployed snapshot returns not-detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "0" "0" || rc=$?
-    [ "$rc" -eq 1 ]
+    if helmrelease_has_remediation_cycle "deployed"; then
+        echo "expected not-detected for deployed-only history" >&2
+        exit 1
+    fi
 }
 
-@test "install zero upgrade empty returns not-detected" {
+@test "deployed then superseded returns not-detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "0" "" || rc=$?
-    [ "$rc" -eq 1 ]
+    statuses=$(printf 'deployed\nsuperseded\n')
+    if helmrelease_has_remediation_cycle "${statuses}"; then
+        echo "expected not-detected for deployed+superseded history" >&2
+        exit 1
+    fi
 }
 
-@test "install empty upgrade zero returns not-detected" {
+@test "single failed snapshot returns detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "" "0" || rc=$?
-    [ "$rc" -eq 1 ]
+    if ! helmrelease_has_remediation_cycle "failed"; then
+        echo "expected detected when history contains failed snapshot" >&2
+        exit 1
+    fi
 }
 
-@test "install one upgrade empty returns detected" {
-    # Canonical race: first install exceeded helm-wait, remediation fired,
-    # no upgrade has happened yet.
+@test "single uninstalled snapshot returns detected" {
+    # The exact signature of the install-remediation race: the first install
+    # exceeded flux's wait budget, remediation uninstalled, the next retry
+    # eventually succeeded. History still carries the uninstalled snapshot.
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "1" "" || rc=$?
-    [ "$rc" -eq 0 ]
+    if ! helmrelease_has_remediation_cycle "uninstalled"; then
+        echo "expected detected when history contains uninstalled snapshot" >&2
+        exit 1
+    fi
 }
 
-@test "install empty upgrade one returns detected" {
+@test "uninstalled then deployed still returns detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "" "1" || rc=$?
-    [ "$rc" -eq 0 ]
+    statuses=$(printf 'uninstalled\ndeployed\n')
+    if ! helmrelease_has_remediation_cycle "${statuses}"; then
+        echo "expected detected despite later successful deploy" >&2
+        exit 1
+    fi
 }
 
-@test "install two upgrade zero returns detected" {
+@test "deployed then failed still returns detected" {
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "2" "0" || rc=$?
-    [ "$rc" -eq 0 ]
+    statuses=$(printf 'deployed\nfailed\n')
+    if ! helmrelease_has_remediation_cycle "${statuses}"; then
+        echo "expected detected when any entry is failed" >&2
+        exit 1
+    fi
 }
 
-@test "install zero upgrade two returns detected" {
-    . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "0" "2" || rc=$?
-    [ "$rc" -eq 0 ]
-}
-
-@test "both counters positive returns detected" {
-    . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "3" "5" || rc=$?
-    [ "$rc" -eq 0 ]
-}
-
-@test "installFailures and upgradeFailures extraction pins HR v2 status shape" {
-    # Pins the Flux HelmRelease v2 status shape that run-kubernetes.sh relies
-    # on. If a future Flux version renames .status.installFailures (or
-    # .status.upgradeFailures), kubectl get -o jsonpath returns an empty
-    # string, the guard quietly says "no cycle", and real remediation loops
-    # slip past the e2e assertion.
-    #
-    # This test uses yq to read the exact path used in the e2e script. yq
-    # evaluates the same json-ish jsonpath against a pinned HR snippet, so
-    # the test fails loudly if the field ever disappears or moves. Cross
-    # reference: vendor/github.com/fluxcd/helm-controller/api/v2/ status
-    # struct field tags.
+@test "installFailures extraction pins HR v2 status.history shape" {
+    # Pins the Flux HelmRelease v2 .status.history[].status shape that
+    # run-kubernetes.sh relies on. If a future flux release renames the
+    # field, the jsonpath returns nothing, the guard reports no cycle,
+    # and real remediation loops slip past the e2e assertion. This test
+    # uses yq to read the exact path used in the e2e script; the upstream
+    # Snapshot type lives at
+    # github.com/fluxcd/helm-controller/api/v2.Snapshot (via go.mod).
     tmp=$(mktemp -d)
     trap 'rm -rf "$tmp"' EXIT
 
@@ -107,25 +98,30 @@ kind: HelmRelease
 metadata:
   name: kubernetes-test
   namespace: tenant-test
-spec:
-  interval: 5m
 status:
-  installFailures: 2
-  upgradeFailures: 0
-  conditions:
-    - type: Ready
-      status: "False"
-      reason: UninstallSucceeded
+  history:
+    - name: kubernetes-test
+      namespace: tenant-test
+      version: 1
+      status: uninstalled
+    - name: kubernetes-test
+      namespace: tenant-test
+      version: 2
+      status: deployed
 YAML
 
-    install_failures=$(yq '.status.installFailures' "$tmp/hr.yaml")
-    upgrade_failures=$(yq '.status.upgradeFailures' "$tmp/hr.yaml")
+    # Default yq output is yaml scalar format, which for string values emits
+    # bare unquoted tokens - matching what kubectl -o jsonpath produces in
+    # e2e. Do not switch to JSON output here; that would quote the values
+    # and break the loop in helmrelease_has_remediation_cycle.
+    statuses=$(yq '.status.history[].status' "$tmp/hr.yaml")
 
-    [ "$install_failures" = "2" ]
-    [ "$upgrade_failures" = "0" ]
+    [ -n "$statuses" ]
+    echo "$statuses" | grep --quiet '^uninstalled$'
 
     . hack/e2e-apps/remediation-guard.sh
-    rc=0
-    helmrelease_has_remediation_cycle "$install_failures" "$upgrade_failures" || rc=$?
-    [ "$rc" -eq 0 ]
+    if ! helmrelease_has_remediation_cycle "$statuses"; then
+        echo "expected detected for pinned HR snippet with uninstalled + deployed history" >&2
+        exit 1
+    fi
 }

From 48312cc3696904235e87fcf2bc33d05e21a77b72 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:29:49 +0300
Subject: [PATCH 19/82] build: wire go-unit-tests into make unit-tests

CI runs make unit-tests on every PR, which already covers helm
unittests and bats (hack/admin-kubeconfig-invariant.bats and
hack/remediation-guard.bats are both picked up by the existing
hack/*.bats glob). What was missing was any go test invocation.

Add a go-unit-tests target scoped to pkg/registry, pkg/config, and
pkg/cmd/server - the cozystack-api surface this repo actually owns
and tests in-tree. Running go test ./... pulls in generated-code
round-trip suites whose behavior is governed by generator tool
versions outside this repo's control; those are better exercised
from their own generator workflows.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 Makefile | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/Makefile b/Makefile
index 0e94c5e6..10a55314 100644
--- a/Makefile
+++ b/Makefile
@@ -82,11 +82,19 @@ test:
 	make -C packages/core/testing apply
 	make -C packages/core/testing test
 
-unit-tests: helm-unit-tests bats-unit-tests
+unit-tests: helm-unit-tests bats-unit-tests go-unit-tests
 
 helm-unit-tests:
 	hack/helm-unit-tests.sh
 
+# Scoped go test over the cozystack-api surface that this repo owns. Kept
+# narrow intentionally - running `go test ./...` pulls in generated code
+# round-trip suites whose behavior depends on tool versions outside this
+# repo's control (kubebuilder, openapi-gen, etc.) and is better exercised
+# from their generator workflows.
+go-unit-tests:
+	go test ./pkg/registry/... ./pkg/config/... ./pkg/cmd/server/...
+
 # Discover every hack/*.bats file that is NOT an e2e test and run it
 # through cozytest.sh. Drop a new *.bats file in hack/ and it is picked
 # up automatically on the next `make unit-tests` run.

From d2e8f7e86cc7157aa83a7130b3f9857f10518218 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:31:29 +0300
Subject: [PATCH 20/82] chore(kubernetes): drop busybox mirror Containerfile,
 pin upstream directly

The image-busybox Makefile target + images/busybox/Dockerfile wrapper
just rebuilt an upstream busybox digest as ghcr.io/cozystack/cozystack/
busybox. Payload is a one-shot sh loop run once per pod; the pinned
upstream digest is already immutable, so maintaining a private mirror
adds churn (rebuild on every release) for no real hardening benefit.
Delete the wrapper and reference docker.io/library/busybox:<digest>
directly from images/busybox.tag. Document the choice in _helpers.tpl.

Also drop the false coupling in the go table test: the "unrelated
kind without configured timeout" case used the real kind name Qdrant,
which tied the test to the Qdrant ApplicationDefinition for no reason.
Switch to a clearly fictional kind so a future Qdrant timeout override
does not break this assertion.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/Makefile                  | 13 +------------
 packages/apps/kubernetes/images/busybox/Dockerfile |  1 -
 packages/apps/kubernetes/templates/_helpers.tpl    |  9 ++++++++-
 pkg/registry/apps/application/rest_timeout_test.go | 10 +++++++---
 4 files changed, 16 insertions(+), 17 deletions(-)
 delete mode 100644 packages/apps/kubernetes/images/busybox/Dockerfile

diff --git a/packages/apps/kubernetes/Makefile b/packages/apps/kubernetes/Makefile
index eb23c2ef..4fea42ef 100644
--- a/packages/apps/kubernetes/Makefile
+++ b/packages/apps/kubernetes/Makefile
@@ -15,7 +15,7 @@ update:
 	hack/update-versions.sh
 	make generate
 
-image: image-ubuntu-container-disk image-kubevirt-cloud-provider image-kubevirt-csi-driver image-cluster-autoscaler image-busybox
+image: image-ubuntu-container-disk image-kubevirt-cloud-provider image-kubevirt-csi-driver image-cluster-autoscaler
 
 image-ubuntu-container-disk:
 	$(foreach ver,$(KUBERNETES_VERSIONS), \
@@ -71,14 +71,3 @@ image-cluster-autoscaler:
 		> images/cluster-autoscaler.tag
 	rm -f images/cluster-autoscaler.json
 
-image-busybox:
-	docker buildx build images/busybox \
-		--tag $(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG)) \
-		--tag $(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG)-$(TAG)) \
-		--cache-from type=registry,ref=$(REGISTRY)/busybox:latest \
-		--cache-to type=inline \
-		--metadata-file images/busybox.json \
-		$(BUILDX_ARGS)
-	echo "$(REGISTRY)/busybox:$(call settag,$(KUBERNETES_PKG_TAG))@$$(yq e '."containerimage.digest"' images/busybox.json -o json -r)" \
-		> images/busybox.tag
-	rm -f images/busybox.json
diff --git a/packages/apps/kubernetes/images/busybox/Dockerfile b/packages/apps/kubernetes/images/busybox/Dockerfile
deleted file mode 100644
index 0c57fd39..00000000
--- a/packages/apps/kubernetes/images/busybox/Dockerfile
+++ /dev/null
@@ -1 +0,0 @@
-FROM docker.io/library/busybox:1.37.0@sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e
diff --git a/packages/apps/kubernetes/templates/_helpers.tpl b/packages/apps/kubernetes/templates/_helpers.tpl
index 89f06934..e6ad9dcf 100644
--- a/packages/apps/kubernetes/templates/_helpers.tpl
+++ b/packages/apps/kubernetes/templates/_helpers.tpl
@@ -60,10 +60,17 @@ exits only when super-admin.svc appears, which happens after kubelet's
 optional-Secret refresh cycle.
 
 The 10m deadline stays strictly below the 15m HelmRelease
-Install.Timeout scoped to the Kubernetes Application kind so the
+Install.Timeout set by cozystack-api for the Kubernetes kind (via the
+release.cozystack.io/helm-install-timeout annotation) so the
 CrashLoopBackOff surfaces before flux remediation fires and uninstalls
 the Cluster CR.
 
+The pinned busybox image in images/busybox.tag points directly at
+docker.io by digest (not mirrored to ghcr.io like the other .tag files
+here): the payload is a one-shot sh loop, the digest pin makes the
+pull immutable, and the cost of maintaining a private mirror of a tiny
+upstream image that does not move often is not worth it.
+
 Call site owns the surrounding volumes block; the kubeconfig volume
 must exist on the pod and mount at /etc/kubernetes/kubeconfig.
 */}}
diff --git a/pkg/registry/apps/application/rest_timeout_test.go b/pkg/registry/apps/application/rest_timeout_test.go
index 8474d1cb..85ac0b97 100644
--- a/pkg/registry/apps/application/rest_timeout_test.go
+++ b/pkg/registry/apps/application/rest_timeout_test.go
@@ -48,9 +48,13 @@ func TestConvertApplicationToHelmRelease_AppliesReleaseConfigTimeout(t *testing.
 			wantSet:    true,
 		},
 		{
-			name:       "Qdrant kind without configured timeout keeps flux defaults",
-			kind:       "Qdrant",
-			prefix:     "qdrant-",
+			// Fictional kind on purpose: the test is about the unset path
+			// regardless of which real Application kind ends up needing a
+			// timeout override. Using a real kind name would create false
+			// coupling to that Application's ApplicationDefinition.
+			name:       "unrelated kind without configured timeout keeps flux defaults",
+			kind:       "PlaceholderKindForDefaults",
+			prefix:     "placeholder-",
 			configured: 0,
 			wantSet:    false,
 		},

From 64b216e10a2392394add9eb7fbf7e908f4c8864a Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:46:01 +0300
Subject: [PATCH 21/82] fix(kubernetes): gate child HelmReleases on tenant etcd
 DataStore

17 child HelmReleases (cilium, coredns, csi, cert-manager,
metrics-server, ...) referenced *-admin-kubeconfig via
kubeConfig.secretRef and rendered even when _namespace.etcd was empty.
On an etcd-less tenant each one sat in NotReady forever because the
admin-kubeconfig Secret only exists after a KamajiControlPlane
reconciles, and KamajiControlPlane now only renders when etcd is set.
The outcome contradicted the "beacon only" contract claimed in the
soft-skip commit.

Extend the existing addon guards to also require _namespace.etcd, and
wrap the four unconditional HelmReleases (csi, metrics-server,
prometheus-operator-crds, volumesnapshot-crd) plus the always-on
cilium/coredns HR resources in the same gate. Add an invariant bats
test that renders the whole chart with etcd empty and asserts zero
HelmReleases reference *-admin-kubeconfig.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/admin-kubeconfig-invariant.bats          | 40 +++++++++++++++++++
 .../helmreleases/cert-manager-crds.yaml       |  2 +-
 .../templates/helmreleases/cert-manager.yaml  |  2 +-
 .../templates/helmreleases/cilium.yaml        |  2 +
 .../templates/helmreleases/coredns.yaml       |  2 +
 .../templates/helmreleases/csi.yaml           |  2 +
 .../templates/helmreleases/fluxcd.yaml        |  2 +-
 .../helmreleases/gateway-api-crds.yaml        |  2 +-
 .../templates/helmreleases/gpu-operator.yaml  |  2 +-
 .../templates/helmreleases/ingress-nginx.yaml |  2 +-
 .../helmreleases/metrics-server.yaml          |  2 +
 .../helmreleases/monitoring-agents.yaml       |  2 +-
 .../prometheus-operator-crds.yaml             |  2 +
 .../templates/helmreleases/velero.yaml        |  2 +-
 .../vertical-pod-autoscaler-crds.yaml         |  2 +-
 .../helmreleases/vertical-pod-autoscaler.yaml |  2 +-
 .../victoria-metrics-operator.yaml            |  2 +-
 .../helmreleases/volumesnapshot-crd.yaml      |  2 +
 18 files changed, 63 insertions(+), 11 deletions(-)

diff --git a/hack/admin-kubeconfig-invariant.bats b/hack/admin-kubeconfig-invariant.bats
index c027e2c4..4c809c41 100644
--- a/hack/admin-kubeconfig-invariant.bats
+++ b/hack/admin-kubeconfig-invariant.bats
@@ -115,3 +115,43 @@
 
   echo "No admin-kubeconfig Deployments rendered for empty etcd (as expected)"
 }
+
+@test "chart emits zero admin-kubeconfig HelmReleases when tenant has no etcd DataStore" {
+  # Same principle as the Deployment variant above, extended to every child
+  # HelmRelease (cilium, coredns, csi, cert-manager, ...). They reference
+  # *-admin-kubeconfig via kubeConfig.secretRef and would otherwise sit in
+  # NotReady forever on an etcd-less tenant, polluting the HelmRelease list
+  # the operator sees and contradicting the "awaiting-etcd beacon only"
+  # contract of the soft-skip path.
+  tmp=$(mktemp -d)
+  trap 'rm -rf "$tmp"' EXIT
+
+  helm template invariant packages/apps/kubernetes \
+    --namespace tenant-root \
+    --set _namespace.etcd="" \
+    --set _namespace.monitoring="" \
+    --set _namespace.ingress="" \
+    --set _namespace.seaweedfs="" \
+    --set _namespace.host="" \
+    --set _cluster.cluster-domain=cozy.local \
+    --set 'nodeGroups=null' \
+    2>/dev/null > "$tmp/rendered.yaml"
+
+  matched=$(
+    yq --output-format=json eval-all '.' "$tmp/rendered.yaml" \
+      | jq -s '
+          map(select(.kind == "HelmRelease")) |
+          map(select(.spec.kubeConfig.secretRef.name | test("-admin-kubeconfig$")?)) |
+          length
+        '
+  )
+
+  if [ "$matched" -ne 0 ]; then
+    echo "Expected zero HelmReleases referencing *-admin-kubeconfig when etcd is empty, got $matched:" >&2
+    yq --output-format=json eval-all '.' "$tmp/rendered.yaml" \
+      | jq -s 'map(select(.kind == "HelmRelease") | .metadata.name)' >&2
+    exit 1
+  fi
+
+  echo "No admin-kubeconfig HelmReleases rendered for empty etcd (as expected)"
+}
diff --git a/packages/apps/kubernetes/templates/helmreleases/cert-manager-crds.yaml b/packages/apps/kubernetes/templates/helmreleases/cert-manager-crds.yaml
index be07a8b9..fd9dac7e 100644
--- a/packages/apps/kubernetes/templates/helmreleases/cert-manager-crds.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/cert-manager-crds.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.certManager.enabled }}
+{{- if and .Values.addons.certManager.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/cert-manager.yaml b/packages/apps/kubernetes/templates/helmreleases/cert-manager.yaml
index 6857581a..700b666e 100644
--- a/packages/apps/kubernetes/templates/helmreleases/cert-manager.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/cert-manager.yaml
@@ -8,7 +8,7 @@ cert-manager:
 {{- end }}
 {{- end }}
 
-{{- if .Values.addons.certManager.enabled }}
+{{- if and .Values.addons.certManager.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/cilium.yaml b/packages/apps/kubernetes/templates/helmreleases/cilium.yaml
index d032f5b6..d8c90cbf 100644
--- a/packages/apps/kubernetes/templates/helmreleases/cilium.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/cilium.yaml
@@ -14,6 +14,7 @@ cilium:
   {{- end }}
 {{- end }}
 
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -55,3 +56,4 @@ spec:
   - name: {{ .Release.Name }}-gateway-api-crds
     namespace: {{ .Release.Namespace }}
   {{- end }}
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/helmreleases/coredns.yaml b/packages/apps/kubernetes/templates/helmreleases/coredns.yaml
index bdb6c682..0711a51d 100644
--- a/packages/apps/kubernetes/templates/helmreleases/coredns.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/coredns.yaml
@@ -4,6 +4,7 @@ coredns:
     clusterIP: "10.95.0.10"
 {{- end }}
 
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -42,3 +43,4 @@ spec:
   {{- end }}
   - name: {{ .Release.Name }}-cilium
     namespace: {{ .Release.Namespace }}
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/helmreleases/csi.yaml b/packages/apps/kubernetes/templates/helmreleases/csi.yaml
index dd2c69a6..109d78e1 100644
--- a/packages/apps/kubernetes/templates/helmreleases/csi.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/csi.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -39,3 +40,4 @@ spec:
   - name: {{ .Release.Name }}
     namespace: {{ .Release.Namespace }}
   {{- end }}
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/helmreleases/fluxcd.yaml b/packages/apps/kubernetes/templates/helmreleases/fluxcd.yaml
index 76499dfe..25fff01c 100644
--- a/packages/apps/kubernetes/templates/helmreleases/fluxcd.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/fluxcd.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.fluxcd.enabled }}
+{{- if and .Values.addons.fluxcd.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/gateway-api-crds.yaml b/packages/apps/kubernetes/templates/helmreleases/gateway-api-crds.yaml
index 2bcc8d4d..b4172ed1 100644
--- a/packages/apps/kubernetes/templates/helmreleases/gateway-api-crds.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/gateway-api-crds.yaml
@@ -1,4 +1,4 @@
-{{- if $.Values.addons.gatewayAPI.enabled }}
+{{- if and $.Values.addons.gatewayAPI.enabled $.Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/gpu-operator.yaml b/packages/apps/kubernetes/templates/helmreleases/gpu-operator.yaml
index 5ef48912..e243bfad 100644
--- a/packages/apps/kubernetes/templates/helmreleases/gpu-operator.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/gpu-operator.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.gpuOperator.enabled }}
+{{- if and .Values.addons.gpuOperator.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/ingress-nginx.yaml b/packages/apps/kubernetes/templates/helmreleases/ingress-nginx.yaml
index 5cafff90..6e2183d3 100644
--- a/packages/apps/kubernetes/templates/helmreleases/ingress-nginx.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/ingress-nginx.yaml
@@ -20,7 +20,7 @@ ingress-nginx:
     node-role.kubernetes.io/ingress-nginx: ""
 {{- end }}
 
-{{- if .Values.addons.ingressNginx.enabled }}
+{{- if and .Values.addons.ingressNginx.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/metrics-server.yaml b/packages/apps/kubernetes/templates/helmreleases/metrics-server.yaml
index 3cc81a14..3e6f9660 100644
--- a/packages/apps/kubernetes/templates/helmreleases/metrics-server.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/metrics-server.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -36,3 +37,4 @@ spec:
     namespace: {{ .Release.Namespace }}
   - name: {{ .Release.Name }}-prometheus-operator-crds
     namespace: {{ .Release.Namespace }}
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/helmreleases/monitoring-agents.yaml b/packages/apps/kubernetes/templates/helmreleases/monitoring-agents.yaml
index ea84dec0..a811f7dd 100644
--- a/packages/apps/kubernetes/templates/helmreleases/monitoring-agents.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/monitoring-agents.yaml
@@ -1,6 +1,6 @@
 {{- $targetTenant := .Values._namespace.monitoring }}
 {{- $clusterDomain := (index .Values._cluster "cluster-domain") | default "cozy.local" }}
-{{- if .Values.addons.monitoringAgents.enabled }}
+{{- if and .Values.addons.monitoringAgents.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/prometheus-operator-crds.yaml b/packages/apps/kubernetes/templates/helmreleases/prometheus-operator-crds.yaml
index 600a7994..3038a058 100644
--- a/packages/apps/kubernetes/templates/helmreleases/prometheus-operator-crds.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/prometheus-operator-crds.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -31,3 +32,4 @@ spec:
   - name: {{ .Release.Name }}
     namespace: {{ .Release.Namespace }}
   {{- end }}
+{{- end }}
diff --git a/packages/apps/kubernetes/templates/helmreleases/velero.yaml b/packages/apps/kubernetes/templates/helmreleases/velero.yaml
index ad236d53..781b9c49 100644
--- a/packages/apps/kubernetes/templates/helmreleases/velero.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/velero.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.velero.enabled }}
+{{- if and .Values.addons.velero.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler-crds.yaml b/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler-crds.yaml
index a3b7a9b4..55a5faac 100644
--- a/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler-crds.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler-crds.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.monitoringAgents.enabled }}
+{{- if and .Values.addons.monitoringAgents.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler.yaml b/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler.yaml
index 178df3e3..74fb5a39 100644
--- a/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/vertical-pod-autoscaler.yaml
@@ -24,7 +24,7 @@ vertical-pod-autoscaler:
         memory: 1600Mi
 {{- end }}
 
-{{- if .Values.addons.monitoringAgents.enabled }}
+{{- if and .Values.addons.monitoringAgents.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/victoria-metrics-operator.yaml b/packages/apps/kubernetes/templates/helmreleases/victoria-metrics-operator.yaml
index 99744277..7302f8f5 100644
--- a/packages/apps/kubernetes/templates/helmreleases/victoria-metrics-operator.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/victoria-metrics-operator.yaml
@@ -1,4 +1,4 @@
-{{- if .Values.addons.monitoringAgents.enabled }}
+{{- if and .Values.addons.monitoringAgents.enabled .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
diff --git a/packages/apps/kubernetes/templates/helmreleases/volumesnapshot-crd.yaml b/packages/apps/kubernetes/templates/helmreleases/volumesnapshot-crd.yaml
index 025f01b7..d50fd93c 100644
--- a/packages/apps/kubernetes/templates/helmreleases/volumesnapshot-crd.yaml
+++ b/packages/apps/kubernetes/templates/helmreleases/volumesnapshot-crd.yaml
@@ -1,3 +1,4 @@
+{{- if .Values._namespace.etcd }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -33,3 +34,4 @@ spec:
   - name: {{ .Release.Name }}
     namespace: {{ .Release.Namespace }}
   {{- end }}
+{{- end }}

From 6072723e1e03c0072fe3bd83e2a5723c50d6ebd8 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:47:19 +0300
Subject: [PATCH 22/82] feat(config): extract + test annotation-timeout parser
 with Flux unit gate
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Pulls the release.cozystack.io/helm-install-timeout parsing out of
start.go into ParseHelmInstallTimeoutAnnotation in pkg/config. The
helper rejects units that time.ParseDuration accepts but Flux
helm-controller rejects (ns, us, µs): feeding one of those through
would cause the HelmRelease admission webhook to reject the object at
install time, giving a silent drop to flux defaults that is hard to
debug. Fail loudly at cozystack-api startup instead.

Adds a table-driven unit test covering: unset (empty), accepted units
ms/s/m/h, compound 2h30m, decimal 1.5m, and the rejected shapes
(bare digits, garbage, negative, ns/us/µs). The test lives in
pkg/config so it runs under the existing go-unit-tests make target.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 pkg/cmd/server/start.go   |  23 ++++----
 pkg/config/config.go      |  42 +++++++++++++-
 pkg/config/config_test.go | 119 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 173 insertions(+), 11 deletions(-)
 create mode 100644 pkg/config/config_test.go

diff --git a/pkg/cmd/server/start.go b/pkg/cmd/server/start.go
index 3e905735..7ebbbc89 100644
--- a/pkg/cmd/server/start.go
+++ b/pkg/cmd/server/start.go
@@ -174,17 +174,20 @@ func (o *CozyServerOptions) Complete() error {
 		// the chart itself depends on (for example, the Kamaji-provisioned
 		// admin-kubeconfig Secret for Kubernetes tenants) need a longer
 		// wait budget than the Flux default. Consumed by the REST storage
-		// layer when building the HelmRelease Spec.
-		if raw, ok := crd.Annotations["release.cozystack.io/helm-install-timeout"]; ok && raw != "" {
-			d, err := time.ParseDuration(raw)
-			if err != nil {
-				return fmt.Errorf(
-					"ApplicationDefinition %q has invalid release.cozystack.io/helm-install-timeout %q: %w",
-					crd.Name, raw, err,
-				)
-			}
-			release.HelmInstallTimeout = d
+		// layer when building the HelmRelease Spec. The parser rejects
+		// units Flux would reject at webhook time, so a bad annotation
+		// surfaces as a loud startup failure instead of a silent drop to
+		// defaults.
+		d, err := config.ParseHelmInstallTimeoutAnnotation(
+			crd.Annotations[config.HelmInstallTimeoutAnnotation],
+		)
+		if err != nil {
+			return fmt.Errorf(
+				"ApplicationDefinition %q has invalid %s annotation: %w",
+				crd.Name, config.HelmInstallTimeoutAnnotation, err,
+			)
 		}
+		release.HelmInstallTimeout = d
 		resource := config.Resource{
 			Application: config.ApplicationConfig{
 				Kind:          crd.Spec.Application.Kind,
diff --git a/pkg/config/config.go b/pkg/config/config.go
index 16cf4f0c..21de27b2 100644
--- a/pkg/config/config.go
+++ b/pkg/config/config.go
@@ -16,7 +16,47 @@ limitations under the License.
 
 package config
 
-import "time"
+import (
+	"fmt"
+	"regexp"
+	"time"
+)
+
+// HelmInstallTimeoutAnnotation is the ApplicationDefinition metadata
+// annotation key that overrides the Flux HelmRelease Install.Timeout and
+// Upgrade.Timeout for a given Application kind.
+const HelmInstallTimeoutAnnotation = "release.cozystack.io/helm-install-timeout"
+
+// helmTimeoutPattern mirrors the CRD validation pattern used by Flux
+// helm-controller on HelmReleaseSpec.Install.Timeout (ms/s/m/h units only).
+// time.ParseDuration accepts ns/us/µs, but Flux rejects them - parsing here
+// with the same shape avoids feeding the controller a value it will later
+// reject at webhook time. See
+// github.com/fluxcd/helm-controller/api/v2 HelmReleaseSpec.Install.Timeout
+// in the go module cache.
+var helmTimeoutPattern = regexp.MustCompile(`^([0-9]+(\.[0-9]+)?(ms|s|m|h))+$`)
+
+// ParseHelmInstallTimeoutAnnotation parses the value of the
+// release.cozystack.io/helm-install-timeout annotation. The empty string is
+// treated as "unset" and returns (0, nil) so callers can leave
+// HelmInstallTimeout zeroed and let flux defaults apply. Values accepted by
+// time.ParseDuration but rejected by Flux (ns/us/µs) return a helpful
+// error instead of silently parsing and failing later at HelmRelease
+// admission.
+func ParseHelmInstallTimeoutAnnotation(raw string) (time.Duration, error) {
+	if raw == "" {
+		return 0, nil
+	}
+	if !helmTimeoutPattern.MatchString(raw) {
+		return 0, fmt.Errorf("must match %s (Flux accepts ms/s/m/h units only), got %q",
+			helmTimeoutPattern, raw)
+	}
+	d, err := time.ParseDuration(raw)
+	if err != nil {
+		return 0, fmt.Errorf("time.ParseDuration(%q): %w", raw, err)
+	}
+	return d, nil
+}
 
 // ResourceConfig represents the structure of the configuration file.
 type ResourceConfig struct {
diff --git a/pkg/config/config_test.go b/pkg/config/config_test.go
new file mode 100644
index 00000000..eb6990af
--- /dev/null
+++ b/pkg/config/config_test.go
@@ -0,0 +1,119 @@
+package config
+
+import (
+	"strings"
+	"testing"
+	"time"
+)
+
+// Cover the annotation parser used by cozystack-api at startup. The parser
+// is consumed by pkg/cmd/server/start.go on every ApplicationDefinition; a
+// typo here silently drops back to flux defaults and the Kubernetes tenant
+// race described in cozystack#2412 reappears, so the table must exercise:
+// - the unset path (empty string treated as "no override"),
+// - every unit Flux accepts (ms, s, m, h),
+// - compound forms (the CRD pattern accepts repeats),
+// - units time.ParseDuration accepts but Flux rejects (ns, us, µs),
+// - outright garbage.
+func TestParseHelmInstallTimeoutAnnotation(t *testing.T) {
+	cases := []struct {
+		name     string
+		input    string
+		want     time.Duration
+		wantErr  bool
+		errMatch string
+	}{
+		{
+			name:  "empty string leaves flux defaults in place",
+			input: "",
+			want:  0,
+		},
+		{
+			name:  "minutes",
+			input: "15m",
+			want:  15 * time.Minute,
+		},
+		{
+			name:  "hours",
+			input: "1h",
+			want:  time.Hour,
+		},
+		{
+			name:  "seconds",
+			input: "45s",
+			want:  45 * time.Second,
+		},
+		{
+			name:  "milliseconds",
+			input: "500ms",
+			want:  500 * time.Millisecond,
+		},
+		{
+			name:  "compound hour and minutes",
+			input: "2h30m",
+			want:  2*time.Hour + 30*time.Minute,
+		},
+		{
+			name:  "decimal minutes",
+			input: "1.5m",
+			want:  90 * time.Second,
+		},
+		{
+			name:     "nanoseconds rejected - Flux CRD pattern excludes ns",
+			input:    "500ns",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+		{
+			name:     "microseconds rejected - Flux CRD pattern excludes us",
+			input:    "500us",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+		{
+			name:     "microseconds unicode rejected",
+			input:    "500µs",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+		{
+			name:     "bare digits rejected",
+			input:    "15",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+		{
+			name:     "garbage rejected",
+			input:    "abc",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+		{
+			name:     "negative rejected",
+			input:    "-15m",
+			wantErr:  true,
+			errMatch: "Flux accepts ms/s/m/h units only",
+		},
+	}
+
+	for _, tc := range cases {
+		t.Run(tc.name, func(t *testing.T) {
+			got, err := ParseHelmInstallTimeoutAnnotation(tc.input)
+			if tc.wantErr {
+				if err == nil {
+					t.Fatalf("expected error, got duration=%v", got)
+				}
+				if tc.errMatch != "" && !strings.Contains(err.Error(), tc.errMatch) {
+					t.Errorf("error %q does not contain %q", err.Error(), tc.errMatch)
+				}
+				return
+			}
+			if err != nil {
+				t.Fatalf("unexpected error: %v", err)
+			}
+			if got != tc.want {
+				t.Errorf("got %v, want %v", got, tc.want)
+			}
+		})
+	}
+}

From b8aec9a9731885cae9217271381680f3e274f46f Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 21:50:10 +0300
Subject: [PATCH 23/82] fix(kubernetes): history guard non-empty check + nits
 from review

- Log .status.history regardless of content so a silently empty result
  (Flux field rename) shows up in CI logs, and treat empty history on
  a Ready HelmRelease as a distinct failure. A Ready HR by definition
  has at least one snapshot; anything else is a shape-drift signal.
- Replace the unquoted heredoc in remediation-guard.sh with a printf |
  grep pipeline. printf %s treats statuses as literal payload (no $
  expansion surprises for future callers), grep --quiet --extended-regexp
  returns exit status the caller can forward directly.
- Share the etcd-absent values file between both invariant tests
  (packages/apps/kubernetes/tests/values-ci-no-etcd.yaml) instead of
  duplicating the --set block.
- Fix typo "override applied" -> "override is applied" in the
  Kubernetes ApplicationDefinition.
- Add a coupling comment in the ApplicationDefinition annotation that
  points at the wait-for-kubeconfig init deadline in _helpers.tpl, so
  a future operator raising the HR timeout updates the init deadline
  too.
- Clarify the per-annotation timeout comment in rest.go so it stops
  implying the feature is Kubernetes-only (it is not - only today's
  one user is).

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/admin-kubeconfig-invariant.bats          | 16 ++------------
 hack/e2e-apps/remediation-guard.sh            | 18 ++++++++--------
 hack/e2e-apps/run-kubernetes.sh               | 12 +++++++++--
 .../kubernetes/tests/values-ci-no-etcd.yaml   |  9 ++++++++
 .../kubernetes-rd/cozyrds/kubernetes.yaml     |  8 ++++++-
 pkg/registry/apps/application/rest.go         | 21 +++++++++++--------
 6 files changed, 49 insertions(+), 35 deletions(-)
 create mode 100644 packages/apps/kubernetes/tests/values-ci-no-etcd.yaml

diff --git a/hack/admin-kubeconfig-invariant.bats b/hack/admin-kubeconfig-invariant.bats
index 4c809c41..a4b98b00 100644
--- a/hack/admin-kubeconfig-invariant.bats
+++ b/hack/admin-kubeconfig-invariant.bats
@@ -85,13 +85,7 @@
 
   helm template invariant packages/apps/kubernetes \
     --namespace tenant-root \
-    --set _namespace.etcd="" \
-    --set _namespace.monitoring="" \
-    --set _namespace.ingress="" \
-    --set _namespace.seaweedfs="" \
-    --set _namespace.host="" \
-    --set _cluster.cluster-domain=cozy.local \
-    --set 'nodeGroups=null' \
+    --values packages/apps/kubernetes/tests/values-ci-no-etcd.yaml \
     2>/dev/null > "$tmp/rendered.yaml"
 
   matched=$(
@@ -128,13 +122,7 @@
 
   helm template invariant packages/apps/kubernetes \
     --namespace tenant-root \
-    --set _namespace.etcd="" \
-    --set _namespace.monitoring="" \
-    --set _namespace.ingress="" \
-    --set _namespace.seaweedfs="" \
-    --set _namespace.host="" \
-    --set _cluster.cluster-domain=cozy.local \
-    --set 'nodeGroups=null' \
+    --values packages/apps/kubernetes/tests/values-ci-no-etcd.yaml \
     2>/dev/null > "$tmp/rendered.yaml"
 
   matched=$(
diff --git a/hack/e2e-apps/remediation-guard.sh b/hack/e2e-apps/remediation-guard.sh
index 38463310..b45561b9 100644
--- a/hack/e2e-apps/remediation-guard.sh
+++ b/hack/e2e-apps/remediation-guard.sh
@@ -26,14 +26,14 @@ helmrelease_has_remediation_cycle() {
     if [ -z "${statuses}" ]; then
         return 1
     fi
-    while IFS= read -r status; do
-        case "${status}" in
-            failed|uninstalled)
-                return 0
-                ;;
-        esac
-    done <<EOF
-${statuses}
-EOF
+    # printf + grep over the pipe, rather than a heredoc plus while read.
+    # printf %s treats the status string as a literal payload, so any stray
+    # $ in a future caller's input does not trigger shell expansion. grep
+    # returns 0 iff at least one line matches the allowlist, which is
+    # exactly the contract the caller wants, so we can return its exit
+    # status directly.
+    if printf '%s\n' "${statuses}" | grep --extended-regexp --quiet '^(failed|uninstalled)$'; then
+        return 0
+    fi
     return 1
 }
diff --git a/hack/e2e-apps/run-kubernetes.sh b/hack/e2e-apps/run-kubernetes.sh
index 64b1ae90..761ee57d 100644
--- a/hack/e2e-apps/run-kubernetes.sh
+++ b/hack/e2e-apps/run-kubernetes.sh
@@ -336,9 +336,17 @@ EOF
   # types are github.com/fluxcd/helm-controller/api v2 Snapshot.
   history_statuses=$(kubectl get hr -n tenant-test "kubernetes-${test_name}" \
     -ojsonpath='{range .status.history[*]}{.status}{"\n"}{end}')
+  # Always emit the raw value so a silent future-Flux field rename shows
+  # up as "empty history on a Ready HR" in CI logs rather than vanishing.
+  echo "Parent HelmRelease history statuses:"
+  printf '%s\n' "${history_statuses:-<empty>}"
+  if [ -z "${history_statuses}" ]; then
+    echo "Unexpected empty .status.history on a Ready HelmRelease - Flux API shape may have changed." >&2
+    kubectl -n tenant-test describe hr "kubernetes-${test_name}" >&2
+    exit 1
+  fi
   if helmrelease_has_remediation_cycle "${history_statuses}"; then
-    echo "Parent HelmRelease entered remediation cycle. History statuses:" >&2
-    printf '%s\n' "${history_statuses}" >&2
+    echo "Parent HelmRelease entered remediation cycle." >&2
     kubectl -n tenant-test describe hr "kubernetes-${test_name}" >&2
     exit 1
   fi
diff --git a/packages/apps/kubernetes/tests/values-ci-no-etcd.yaml b/packages/apps/kubernetes/tests/values-ci-no-etcd.yaml
new file mode 100644
index 00000000..c7c8196f
--- /dev/null
+++ b/packages/apps/kubernetes/tests/values-ci-no-etcd.yaml
@@ -0,0 +1,9 @@
+_namespace:
+  etcd: ""
+  monitoring: ""
+  ingress: ""
+  seaweedfs: ""
+  host: ""
+_cluster:
+  cluster-domain: cozy.local
+nodeGroups: null
diff --git a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
index 1128abe8..aaa01247 100644
--- a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
+++ b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
@@ -7,8 +7,14 @@ metadata:
     # Secret is provisioned asynchronously. Cold Kamaji start (image pull +
     # etcd + apiserver Ready) plus admin-kubeconfig generation can exceed
     # Flux helm-controller's default wait budget, causing remediation loops
-    # that uninstall the Cluster CR. This override applied by cozystack-api
+    # that uninstall the Cluster CR. This override is applied by cozystack-api
     # to the HelmRelease Spec.Install.Timeout and Spec.Upgrade.Timeout.
+    #
+    # Coupling: the wait-for-kubeconfig init container in
+    # packages/apps/kubernetes/templates/_helpers.tpl hard-codes a 10m
+    # deadline chosen to stay strictly below this value so the pod's
+    # CrashLoopBackOff surfaces before flux remediation fires. If this
+    # annotation is raised, update that init deadline correspondingly.
     release.cozystack.io/helm-install-timeout: "15m"
 spec:
   application:
diff --git a/pkg/registry/apps/application/rest.go b/pkg/registry/apps/application/rest.go
index 406d3738..d233ecc3 100644
--- a/pkg/registry/apps/application/rest.go
+++ b/pkg/registry/apps/application/rest.go
@@ -1528,15 +1528,18 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (*
 		},
 	}
 
-	// Per-Application HelmRelease wait budget. When an ApplicationDefinition
-	// sets release.cozystack.io/helm-install-timeout, the annotation is
-	// parsed at startup into ReleaseConfig.HelmInstallTimeout and applied
-	// to both Install and Upgrade here. Applications that leave it unset
-	// (the common case) keep flux defaults, so their failed installs
-	// remediate on the normal cadence. Needed for the Kubernetes kind
-	// because its parent chart contains CAPI/Kamaji resources whose
-	// admin-kubeconfig Secret is provisioned asynchronously and Kamaji
-	// cold-start routinely exceeds flux's default wait budget.
+	// Per-Application HelmRelease wait budget. The mechanism is generic:
+	// an ApplicationDefinition that sets
+	// release.cozystack.io/helm-install-timeout gets Install.Timeout and
+	// Upgrade.Timeout populated from ReleaseConfig.HelmInstallTimeout
+	// (parsed at startup). Applications that leave it unset keep flux
+	// defaults so their failed installs remediate on the normal cadence.
+	// Today only kubernetes-rd carries the annotation because the
+	// Kubernetes Application's parent chart contains CAPI/Kamaji
+	// resources whose admin-kubeconfig Secret is provisioned
+	// asynchronously and Kamaji cold-start routinely exceeds flux's
+	// default wait budget; any future kind with the same shape can opt
+	// in by setting the same annotation.
 	if r.releaseConfig.HelmInstallTimeout > 0 {
 		timeout := metav1.Duration{Duration: r.releaseConfig.HelmInstallTimeout}
 		helmRelease.Spec.Install.Timeout = &timeout

From 39b8f0252b6fb07d510c83aacf48fc20d7fdd96c Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 16 Apr 2026 22:32:24 +0300
Subject: [PATCH 24/82] test(hack): rename remediation-guard bats test to match
 what it pins

The test body asserts .status.history[].status extraction, but the
test name still referenced the old installFailures counter (leftover
from when the guard used that field before switching to status.history
to avoid ClearFailures zeroing the counters on successful reconcile).

Address review feedback from coderabbitai on
hack/remediation-guard.bats:84: rename so grep for what the test
actually pins matches the test name.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 hack/remediation-guard.bats | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/hack/remediation-guard.bats b/hack/remediation-guard.bats
index 30e4e04c..092fe06d 100644
--- a/hack/remediation-guard.bats
+++ b/hack/remediation-guard.bats
@@ -81,7 +81,7 @@
     fi
 }
 
-@test "installFailures extraction pins HR v2 status.history shape" {
+@test "status.history extraction pins HR v2 status.history shape" {
     # Pins the Flux HelmRelease v2 .status.history[].status shape that
     # run-kubernetes.sh relies on. If a future flux release renames the
     # field, the jsonpath returns nothing, the guard reports no cycle,

From e148343fd9bf9440d726bf01ebb07374119927d0 Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Sun, 19 Apr 2026 19:01:53 +0300
Subject: [PATCH 25/82] fix(kamaji): increase memory limits and add startup
 probe
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Increase memory limit from 500Mi to 512Mi
- Increase memory request from 100Mi to 256Mi
- Add startup probe with 60s timeout (12 attempts × 5s)
- Increase readiness/liveness initialDelaySeconds from 5/15 to 30s

This fixes OOMKilled crashes observed in production where kamaji
controller was being killed due to insufficient memory during startup.

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 packages/system/kamaji/values.yaml | 27 +++++++++++++++++++++++++--
 1 file changed, 25 insertions(+), 2 deletions(-)

diff --git a/packages/system/kamaji/values.yaml b/packages/system/kamaji/values.yaml
index b1ca3cb7..e6dbe99b 100644
--- a/packages/system/kamaji/values.yaml
+++ b/packages/system/kamaji/values.yaml
@@ -8,9 +8,32 @@ kamaji:
   resources:
     limits:
       cpu: 200m
-      memory: 500Mi
+      memory: 512Mi
     requests:
       cpu: 100m
-      memory: 100Mi
+      memory: 256Mi
+  startupProbe:
+    httpGet:
+      path: /healthz
+      port: healthcheck
+    initialDelaySeconds: 5
+    periodSeconds: 5
+    timeoutSeconds: 3
+    successThreshold: 1
+    failureThreshold: 12
+  livenessProbe:
+    httpGet:
+      path: /healthz
+      port: healthcheck
+    initialDelaySeconds: 30
+    periodSeconds: 20
+    timeoutSeconds: 1
+  readinessProbe:
+    httpGet:
+      path: /readyz
+      port: healthcheck
+    initialDelaySeconds: 30
+    periodSeconds: 10
+    timeoutSeconds: 1
   extraArgs:
     - --migrate-image=ghcr.io/cozystack/cozystack/kamaji:v1.3.0-rc.1@sha256:8c9af083b60600c0800eb56a2cda75f26007b7272a1cf019140de003bfce1a4d

From 531bc00524be0789ba99e2082234a1104e010b51 Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Wed, 8 Apr 2026 22:17:34 +0300
Subject: [PATCH 26/82] feat(postgres): add serverName parameter for backup
 recovery

Add serverName field to bootstrap configuration to explicitly specify
Barman server name from backup.info. This fixes "no target backup found"
errors when server_name in backup.info differs from Kubernetes cluster name.

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 api/apps/v1alpha1/postgresql/types.go             |  5 ++++-
 packages/apps/postgres/README.md                  | 13 +++++++------
 packages/apps/postgres/templates/db.yaml          |  3 +++
 packages/apps/postgres/values.schema.json         |  7 ++++++-
 packages/apps/postgres/values.yaml                |  4 +++-
 packages/system/postgres-rd/cozyrds/postgres.yaml |  4 ++--
 6 files changed, 25 insertions(+), 11 deletions(-)

diff --git a/api/apps/v1alpha1/postgresql/types.go b/api/apps/v1alpha1/postgresql/types.go
index fa85d51f..56580018 100644
--- a/api/apps/v1alpha1/postgresql/types.go
+++ b/api/apps/v1alpha1/postgresql/types.go
@@ -86,12 +86,15 @@ type Bootstrap struct {
 	// Whether to restore from a backup.
 	// +kubebuilder:default:=false
 	Enabled bool `json:"enabled"`
-	// Previous cluster name before deletion.
+	// Previous cluster name before deletion (matches serverName in backup.info).
 	// +kubebuilder:default:=""
 	OldName string `json:"oldName"`
 	// Timestamp (RFC3339) for point-in-time recovery; empty means latest.
 	// +kubebuilder:default:=""
 	RecoveryTime string `json:"recoveryTime,omitempty"`
+	// Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.
+	// +kubebuilder:default:=""
+	ServerName string `json:"serverName,omitempty"`
 }
 
 type Database struct {
diff --git a/packages/apps/postgres/README.md b/packages/apps/postgres/README.md
index 5a550f7a..648b3d71 100644
--- a/packages/apps/postgres/README.md
+++ b/packages/apps/postgres/README.md
@@ -133,12 +133,13 @@ See:
 
 ### Bootstrap (recovery) parameters
 
-| Name                     | Description                                                         | Type     | Value   |
-| ------------------------ | ------------------------------------------------------------------- | -------- | ------- |
-| `bootstrap`              | Bootstrap configuration.                                            | `object` | `{}`    |
-| `bootstrap.enabled`      | Whether to restore from a backup.                                   | `bool`   | `false` |
-| `bootstrap.recoveryTime` | Timestamp (RFC3339) for point-in-time recovery; empty means latest. | `string` | `""`    |
-| `bootstrap.oldName`      | Previous cluster name before deletion.                              | `string` | `""`    |
+| Name                     | Description                                                                                                                                        | Type     | Value   |
+| ------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------- | -------- | ------- |
+| `bootstrap`              | Bootstrap configuration.                                                                                                                           | `object` | `{}`    |
+| `bootstrap.enabled`      | Whether to restore from a backup.                                                                                                                  | `bool`   | `false` |
+| `bootstrap.recoveryTime` | Timestamp (RFC3339) for point-in-time recovery; empty means latest.                                                                                | `string` | `""`    |
+| `bootstrap.oldName`      | Previous cluster name before deletion (matches serverName in backup.info).                                                                         | `string` | `""`    |
+| `bootstrap.serverName`   | Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name. | `string` | `""`    |
 
 
 ## Parameter examples and reference
diff --git a/packages/apps/postgres/templates/db.yaml b/packages/apps/postgres/templates/db.yaml
index 5c40b2c7..2cdaec1d 100644
--- a/packages/apps/postgres/templates/db.yaml
+++ b/packages/apps/postgres/templates/db.yaml
@@ -32,6 +32,9 @@ spec:
     - name: {{ .Values.bootstrap.oldName }}
       barmanObjectStore:
         destinationPath: {{ .Values.backup.destinationPath }}
+        {{- if .Values.bootstrap.serverName }}
+        serverName: {{ .Values.bootstrap.serverName }}
+        {{- end }}
         endpointURL: {{ .Values.backup.endpointURL }}
         s3Credentials:
           accessKeyId:
diff --git a/packages/apps/postgres/values.schema.json b/packages/apps/postgres/values.schema.json
index b2a4aeba..28acc5a3 100644
--- a/packages/apps/postgres/values.schema.json
+++ b/packages/apps/postgres/values.schema.json
@@ -246,7 +246,7 @@
           "default": false
         },
         "oldName": {
-          "description": "Previous cluster name before deletion.",
+          "description": "Previous cluster name before deletion (matches serverName in backup.info).",
           "type": "string",
           "default": ""
         },
@@ -254,6 +254,11 @@
           "description": "Timestamp (RFC3339) for point-in-time recovery; empty means latest.",
           "type": "string",
           "default": ""
+        },
+        "serverName": {
+          "description": "Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.",
+          "type": "string",
+          "default": ""
         }
       }
     }
diff --git a/packages/apps/postgres/values.yaml b/packages/apps/postgres/values.yaml
index b8f07f63..a4a4bf86 100644
--- a/packages/apps/postgres/values.yaml
+++ b/packages/apps/postgres/values.yaml
@@ -153,7 +153,8 @@ backup:
 ## @typedef {struct} Bootstrap - Bootstrap configuration for restoring a database cluster from a backup.
 ## @field {bool} enabled - Whether to restore from a backup.
 ## @field {string} [recoveryTime] - Timestamp (RFC3339) for point-in-time recovery; empty means latest.
-## @field {string} oldName - Previous cluster name before deletion.
+## @field {string} oldName - Previous cluster name before deletion (matches serverName in backup.info).
+## @field {string} [serverName] - Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.
 
 ## @param {Bootstrap} bootstrap - Bootstrap configuration.
 bootstrap:
@@ -161,3 +162,4 @@ bootstrap:
   # example: 2020-11-26 15:22:00.00000+00
   recoveryTime: ""
   oldName: ""
+  serverName: ""
diff --git a/packages/system/postgres-rd/cozyrds/postgres.yaml b/packages/system/postgres-rd/cozyrds/postgres.yaml
index c74c5783..41457db4 100644
--- a/packages/system/postgres-rd/cozyrds/postgres.yaml
+++ b/packages/system/postgres-rd/cozyrds/postgres.yaml
@@ -8,7 +8,7 @@ spec:
     singular: postgres
     plural: postgreses
     openAPISchema: |-
-      {"title":"Chart Values","type":"object","properties":{"replicas":{"description":"Number of Postgres replicas.","type":"integer","default":2},"resources":{"description":"Explicit CPU and memory configuration for each PostgreSQL replica. When omitted, the preset defined in `resourcesPreset` is applied.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Default sizing preset used when `resources` is omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]},"size":{"description":"Persistent Volume Claim size available for application data.","default":"10Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":""},"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"version":{"description":"PostgreSQL major version to deploy","type":"string","default":"v18","enum":["v18","v17","v16","v15","v14","v13"]},"postgresql":{"description":"PostgreSQL server configuration.","type":"object","default":{},"properties":{"parameters":{"description":"PostgreSQL server parameters.","type":"object","default":{},"properties":{"max_connections":{"description":"Maximum number of concurrent connections to the database server.","type":"integer","default":100}}}}},"quorum":{"description":"Quorum configuration for synchronous replication.","type":"object","default":{},"required":["maxSyncReplicas","minSyncReplicas"],"properties":{"maxSyncReplicas":{"description":"Maximum number of synchronous replicas allowed (must be less than total replicas).","type":"integer","default":0},"minSyncReplicas":{"description":"Minimum number of synchronous replicas required for commit.","type":"integer","default":0}}},"users":{"description":"Users configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"password":{"description":"Password for the user.","type":"string"},"replication":{"description":"Whether the user has replication privileges.","type":"boolean"}}}},"databases":{"description":"Databases configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"extensions":{"description":"List of enabled PostgreSQL extensions.","type":"array","items":{"type":"string"}},"roles":{"description":"Roles assigned to users.","type":"object","properties":{"admin":{"description":"List of users with admin privileges.","type":"array","items":{"type":"string"}},"readonly":{"description":"List of users with read-only privileges.","type":"array","items":{"type":"string"}}}}}}},"backup":{"description":"Backup configuration.","type":"object","default":{},"required":["enabled"],"properties":{"destinationPath":{"description":"Destination path for backups (e.g. s3://bucket/path/).","type":"string","default":"s3://bucket/path/to/folder/"},"enabled":{"description":"Enable regular backups.","type":"boolean","default":false},"endpointURL":{"description":"S3 endpoint URL for uploads.","type":"string","default":"http://minio-gateway-service:9000"},"retentionPolicy":{"description":"Retention policy (e.g. \"30d\").","type":"string","default":"30d"},"s3AccessKey":{"description":"Access key for S3 authentication.","type":"string","default":"<your-access-key>"},"s3SecretKey":{"description":"Secret key for S3 authentication.","type":"string","default":"<your-secret-key>"},"schedule":{"description":"Cron schedule for automated backups.","type":"string","default":"0 2 * * * *"}}},"bootstrap":{"description":"Bootstrap configuration.","type":"object","default":{},"required":["enabled","oldName"],"properties":{"enabled":{"description":"Whether to restore from a backup.","type":"boolean","default":false},"oldName":{"description":"Previous cluster name before deletion.","type":"string","default":""},"recoveryTime":{"description":"Timestamp (RFC3339) for point-in-time recovery; empty means latest.","type":"string","default":""}}}}}
+      {"title":"Chart Values","type":"object","properties":{"replicas":{"description":"Number of Postgres replicas.","type":"integer","default":2},"resources":{"description":"Explicit CPU and memory configuration for each PostgreSQL replica. When omitted, the preset defined in `resourcesPreset` is applied.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Default sizing preset used when `resources` is omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]},"size":{"description":"Persistent Volume Claim size available for application data.","default":"10Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":""},"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"version":{"description":"PostgreSQL major version to deploy","type":"string","default":"v18","enum":["v18","v17","v16","v15","v14","v13"]},"postgresql":{"description":"PostgreSQL server configuration.","type":"object","default":{},"properties":{"parameters":{"description":"PostgreSQL server parameters.","type":"object","default":{},"properties":{"max_connections":{"description":"Maximum number of concurrent connections to the database server.","type":"integer","default":100}}}}},"quorum":{"description":"Quorum configuration for synchronous replication.","type":"object","default":{},"required":["maxSyncReplicas","minSyncReplicas"],"properties":{"maxSyncReplicas":{"description":"Maximum number of synchronous replicas allowed (must be less than total replicas).","type":"integer","default":0},"minSyncReplicas":{"description":"Minimum number of synchronous replicas required for commit.","type":"integer","default":0}}},"users":{"description":"Users configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"password":{"description":"Password for the user.","type":"string"},"replication":{"description":"Whether the user has replication privileges.","type":"boolean"}}}},"databases":{"description":"Databases configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"extensions":{"description":"List of enabled PostgreSQL extensions.","type":"array","items":{"type":"string"}},"roles":{"description":"Roles assigned to users.","type":"object","properties":{"admin":{"description":"List of users with admin privileges.","type":"array","items":{"type":"string"}},"readonly":{"description":"List of users with read-only privileges.","type":"array","items":{"type":"string"}}}}}}},"backup":{"description":"Backup configuration.","type":"object","default":{},"required":["enabled"],"properties":{"destinationPath":{"description":"Destination path for backups (e.g. s3://bucket/path/).","type":"string","default":"s3://bucket/path/to/folder/"},"enabled":{"description":"Enable regular backups.","type":"boolean","default":false},"endpointURL":{"description":"S3 endpoint URL for uploads.","type":"string","default":"http://minio-gateway-service:9000"},"retentionPolicy":{"description":"Retention policy (e.g. \"30d\").","type":"string","default":"30d"},"s3AccessKey":{"description":"Access key for S3 authentication.","type":"string","default":"<your-access-key>"},"s3SecretKey":{"description":"Secret key for S3 authentication.","type":"string","default":"<your-secret-key>"},"schedule":{"description":"Cron schedule for automated backups.","type":"string","default":"0 2 * * * *"}}},"bootstrap":{"description":"Bootstrap configuration.","type":"object","default":{},"required":["enabled","oldName"],"properties":{"enabled":{"description":"Whether to restore from a backup.","type":"boolean","default":false},"oldName":{"description":"Previous cluster name before deletion (matches serverName in backup.info).","type":"string","default":""},"recoveryTime":{"description":"Timestamp (RFC3339) for point-in-time recovery; empty means latest.","type":"string","default":""},"serverName":{"description":"Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.","type":"string","default":""}}}}}
   release:
     prefix: postgres-
     labels:
@@ -33,7 +33,7 @@ spec:
     #    labelSelector:
     #      helm.toolkit.fluxcd.io/name: "{reqs[0]['metadata','name']}"
 
-    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "replicas"], ["spec", "resources"], ["spec", "resourcesPreset"], ["spec", "size"], ["spec", "storageClass"], ["spec", "external"], ["spec", "version"], ["spec", "postgresql"], ["spec", "postgresql", "parameters"], ["spec", "postgresql", "parameters", "max_connections"], ["spec", "quorum"], ["spec", "quorum", "minSyncReplicas"], ["spec", "quorum", "maxSyncReplicas"], ["spec", "users"], ["spec", "databases"], ["spec", "backup"], ["spec", "backup", "enabled"], ["spec", "backup", "retentionPolicy"], ["spec", "backup", "destinationPath"], ["spec", "backup", "endpointURL"], ["spec", "backup", "schedule"], ["spec", "backup", "s3AccessKey"], ["spec", "backup", "s3SecretKey"], ["spec", "bootstrap"], ["spec", "bootstrap", "enabled"], ["spec", "bootstrap", "recoveryTime"], ["spec", "bootstrap", "oldName"]]
+    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "replicas"], ["spec", "resources"], ["spec", "resourcesPreset"], ["spec", "size"], ["spec", "storageClass"], ["spec", "external"], ["spec", "version"], ["spec", "postgresql"], ["spec", "postgresql", "parameters"], ["spec", "postgresql", "parameters", "max_connections"], ["spec", "quorum"], ["spec", "quorum", "minSyncReplicas"], ["spec", "quorum", "maxSyncReplicas"], ["spec", "users"], ["spec", "databases"], ["spec", "backup"], ["spec", "backup", "enabled"], ["spec", "backup", "retentionPolicy"], ["spec", "backup", "destinationPath"], ["spec", "backup", "endpointURL"], ["spec", "backup", "schedule"], ["spec", "backup", "s3AccessKey"], ["spec", "backup", "s3SecretKey"], ["spec", "bootstrap"], ["spec", "bootstrap", "enabled"], ["spec", "bootstrap", "recoveryTime"], ["spec", "bootstrap", "oldName"], ["spec", "bootstrap", "serverName"]]
   secrets:
     exclude: []
     include:

From 9f41dc3228b18308f9cd6819a8b2470778f27f0c Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Tue, 21 Apr 2026 13:17:32 +0300
Subject: [PATCH 27/82] docs(postgres): clarify bootstrap field descriptions

Update oldName and serverName field descriptions based on code review
feedback to avoid confusion about their actual roles:

- oldName: Remove misleading "(matches serverName in backup.info)"
  text. This field represents the Kubernetes cluster resource name,
  not the Barman server name.

- serverName: Provide clearer explanation that it's the S3 path prefix
  (barmanObjectStore.serverName) used by the original cluster, and should
  only be set when it differs from the Kubernetes resource name.

Updated in:
- values.yaml (source of truth for field documentation)
- types.go (Go API type comments)
- values.schema.json (JSON schema for validation)
- postgres.yaml (CRD with embedded OpenAPI schema)

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 api/apps/v1alpha1/postgresql/types.go             |  4 ++--
 api/backups/v1alpha1/zz_generated.deepcopy.go     |  5 +++++
 packages/apps/postgres/README.md                  | 14 +++++++-------
 packages/apps/postgres/values.schema.json         |  4 ++--
 packages/apps/postgres/values.yaml                |  4 ++--
 packages/system/postgres-rd/cozyrds/postgres.yaml |  2 +-
 6 files changed, 19 insertions(+), 14 deletions(-)

diff --git a/api/apps/v1alpha1/postgresql/types.go b/api/apps/v1alpha1/postgresql/types.go
index 56580018..a2ff77a8 100644
--- a/api/apps/v1alpha1/postgresql/types.go
+++ b/api/apps/v1alpha1/postgresql/types.go
@@ -86,13 +86,13 @@ type Bootstrap struct {
 	// Whether to restore from a backup.
 	// +kubebuilder:default:=false
 	Enabled bool `json:"enabled"`
-	// Previous cluster name before deletion (matches serverName in backup.info).
+	// Previous cluster name before deletion.
 	// +kubebuilder:default:=""
 	OldName string `json:"oldName"`
 	// Timestamp (RFC3339) for point-in-time recovery; empty means latest.
 	// +kubebuilder:default:=""
 	RecoveryTime string `json:"recoveryTime,omitempty"`
-	// Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.
+	// Barman server name (S3 path prefix) used by the original cluster when writing backups. Set this only when the original cluster had an explicit barmanObjectStore.serverName that differed from its Kubernetes resource name.
 	// +kubebuilder:default:=""
 	ServerName string `json:"serverName,omitempty"`
 }
diff --git a/api/backups/v1alpha1/zz_generated.deepcopy.go b/api/backups/v1alpha1/zz_generated.deepcopy.go
index 89f6171f..61b8f839 100644
--- a/api/backups/v1alpha1/zz_generated.deepcopy.go
+++ b/api/backups/v1alpha1/zz_generated.deepcopy.go
@@ -620,6 +620,11 @@ func (in *RestoreJobSpec) DeepCopyInto(out *RestoreJobSpec) {
 		*out = new(v1.TypedLocalObjectReference)
 		(*in).DeepCopyInto(*out)
 	}
+	if in.Options != nil {
+		in, out := &in.Options, &out.Options
+		*out = new(runtime.RawExtension)
+		(*in).DeepCopyInto(*out)
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new RestoreJobSpec.
diff --git a/packages/apps/postgres/README.md b/packages/apps/postgres/README.md
index 648b3d71..4cda7284 100644
--- a/packages/apps/postgres/README.md
+++ b/packages/apps/postgres/README.md
@@ -133,13 +133,13 @@ See:
 
 ### Bootstrap (recovery) parameters
 
-| Name                     | Description                                                                                                                                        | Type     | Value   |
-| ------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------- | -------- | ------- |
-| `bootstrap`              | Bootstrap configuration.                                                                                                                           | `object` | `{}`    |
-| `bootstrap.enabled`      | Whether to restore from a backup.                                                                                                                  | `bool`   | `false` |
-| `bootstrap.recoveryTime` | Timestamp (RFC3339) for point-in-time recovery; empty means latest.                                                                                | `string` | `""`    |
-| `bootstrap.oldName`      | Previous cluster name before deletion (matches serverName in backup.info).                                                                         | `string` | `""`    |
-| `bootstrap.serverName`   | Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name. | `string` | `""`    |
+| Name                     | Description                                                                                                                                                                                                                  | Type     | Value   |
+| ------------------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------- | ------- |
+| `bootstrap`              | Bootstrap configuration.                                                                                                                                                                                                     | `object` | `{}`    |
+| `bootstrap.enabled`      | Whether to restore from a backup.                                                                                                                                                                                            | `bool`   | `false` |
+| `bootstrap.recoveryTime` | Timestamp (RFC3339) for point-in-time recovery; empty means latest.                                                                                                                                                          | `string` | `""`    |
+| `bootstrap.oldName`      | Previous cluster name before deletion.                                                                                                                                                                                       | `string` | `""`    |
+| `bootstrap.serverName`   | Barman server name (S3 path prefix) used by the original cluster when writing backups. Set this only when the original cluster had an explicit barmanObjectStore.serverName that differed from its Kubernetes resource name. | `string` | `""`    |
 
 
 ## Parameter examples and reference
diff --git a/packages/apps/postgres/values.schema.json b/packages/apps/postgres/values.schema.json
index 28acc5a3..98e29822 100644
--- a/packages/apps/postgres/values.schema.json
+++ b/packages/apps/postgres/values.schema.json
@@ -246,7 +246,7 @@
           "default": false
         },
         "oldName": {
-          "description": "Previous cluster name before deletion (matches serverName in backup.info).",
+          "description": "Previous cluster name before deletion.",
           "type": "string",
           "default": ""
         },
@@ -256,7 +256,7 @@
           "default": ""
         },
         "serverName": {
-          "description": "Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.",
+          "description": "Barman server name (S3 path prefix) used by the original cluster when writing backups. Set this only when the original cluster had an explicit barmanObjectStore.serverName that differed from its Kubernetes resource name.",
           "type": "string",
           "default": ""
         }
diff --git a/packages/apps/postgres/values.yaml b/packages/apps/postgres/values.yaml
index a4a4bf86..2ceaa9cf 100644
--- a/packages/apps/postgres/values.yaml
+++ b/packages/apps/postgres/values.yaml
@@ -153,8 +153,8 @@ backup:
 ## @typedef {struct} Bootstrap - Bootstrap configuration for restoring a database cluster from a backup.
 ## @field {bool} enabled - Whether to restore from a backup.
 ## @field {string} [recoveryTime] - Timestamp (RFC3339) for point-in-time recovery; empty means latest.
-## @field {string} oldName - Previous cluster name before deletion (matches serverName in backup.info).
-## @field {string} [serverName] - Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.
+## @field {string} oldName - Previous cluster name before deletion.
+## @field {string} [serverName] - Barman server name (S3 path prefix) used by the original cluster when writing backups. Set this only when the original cluster had an explicit barmanObjectStore.serverName that differed from its Kubernetes resource name.
 
 ## @param {Bootstrap} bootstrap - Bootstrap configuration.
 bootstrap:
diff --git a/packages/system/postgres-rd/cozyrds/postgres.yaml b/packages/system/postgres-rd/cozyrds/postgres.yaml
index 41457db4..72d013f1 100644
--- a/packages/system/postgres-rd/cozyrds/postgres.yaml
+++ b/packages/system/postgres-rd/cozyrds/postgres.yaml
@@ -8,7 +8,7 @@ spec:
     singular: postgres
     plural: postgreses
     openAPISchema: |-
-      {"title":"Chart Values","type":"object","properties":{"replicas":{"description":"Number of Postgres replicas.","type":"integer","default":2},"resources":{"description":"Explicit CPU and memory configuration for each PostgreSQL replica. When omitted, the preset defined in `resourcesPreset` is applied.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Default sizing preset used when `resources` is omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]},"size":{"description":"Persistent Volume Claim size available for application data.","default":"10Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":""},"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"version":{"description":"PostgreSQL major version to deploy","type":"string","default":"v18","enum":["v18","v17","v16","v15","v14","v13"]},"postgresql":{"description":"PostgreSQL server configuration.","type":"object","default":{},"properties":{"parameters":{"description":"PostgreSQL server parameters.","type":"object","default":{},"properties":{"max_connections":{"description":"Maximum number of concurrent connections to the database server.","type":"integer","default":100}}}}},"quorum":{"description":"Quorum configuration for synchronous replication.","type":"object","default":{},"required":["maxSyncReplicas","minSyncReplicas"],"properties":{"maxSyncReplicas":{"description":"Maximum number of synchronous replicas allowed (must be less than total replicas).","type":"integer","default":0},"minSyncReplicas":{"description":"Minimum number of synchronous replicas required for commit.","type":"integer","default":0}}},"users":{"description":"Users configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"password":{"description":"Password for the user.","type":"string"},"replication":{"description":"Whether the user has replication privileges.","type":"boolean"}}}},"databases":{"description":"Databases configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"extensions":{"description":"List of enabled PostgreSQL extensions.","type":"array","items":{"type":"string"}},"roles":{"description":"Roles assigned to users.","type":"object","properties":{"admin":{"description":"List of users with admin privileges.","type":"array","items":{"type":"string"}},"readonly":{"description":"List of users with read-only privileges.","type":"array","items":{"type":"string"}}}}}}},"backup":{"description":"Backup configuration.","type":"object","default":{},"required":["enabled"],"properties":{"destinationPath":{"description":"Destination path for backups (e.g. s3://bucket/path/).","type":"string","default":"s3://bucket/path/to/folder/"},"enabled":{"description":"Enable regular backups.","type":"boolean","default":false},"endpointURL":{"description":"S3 endpoint URL for uploads.","type":"string","default":"http://minio-gateway-service:9000"},"retentionPolicy":{"description":"Retention policy (e.g. \"30d\").","type":"string","default":"30d"},"s3AccessKey":{"description":"Access key for S3 authentication.","type":"string","default":"<your-access-key>"},"s3SecretKey":{"description":"Secret key for S3 authentication.","type":"string","default":"<your-secret-key>"},"schedule":{"description":"Cron schedule for automated backups.","type":"string","default":"0 2 * * * *"}}},"bootstrap":{"description":"Bootstrap configuration.","type":"object","default":{},"required":["enabled","oldName"],"properties":{"enabled":{"description":"Whether to restore from a backup.","type":"boolean","default":false},"oldName":{"description":"Previous cluster name before deletion (matches serverName in backup.info).","type":"string","default":""},"recoveryTime":{"description":"Timestamp (RFC3339) for point-in-time recovery; empty means latest.","type":"string","default":""},"serverName":{"description":"Barman server name from the old cluster's backup.info. Use when the original cluster used a different serverName than its Kubernetes cluster name.","type":"string","default":""}}}}}
+      {"title":"Chart Values","type":"object","properties":{"replicas":{"description":"Number of Postgres replicas.","type":"integer","default":2},"resources":{"description":"Explicit CPU and memory configuration for each PostgreSQL replica. When omitted, the preset defined in `resourcesPreset` is applied.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available to each replica.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Default sizing preset used when `resources` is omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]},"size":{"description":"Persistent Volume Claim size available for application data.","default":"10Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":""},"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"version":{"description":"PostgreSQL major version to deploy","type":"string","default":"v18","enum":["v18","v17","v16","v15","v14","v13"]},"postgresql":{"description":"PostgreSQL server configuration.","type":"object","default":{},"properties":{"parameters":{"description":"PostgreSQL server parameters.","type":"object","default":{},"properties":{"max_connections":{"description":"Maximum number of concurrent connections to the database server.","type":"integer","default":100}}}}},"quorum":{"description":"Quorum configuration for synchronous replication.","type":"object","default":{},"required":["maxSyncReplicas","minSyncReplicas"],"properties":{"maxSyncReplicas":{"description":"Maximum number of synchronous replicas allowed (must be less than total replicas).","type":"integer","default":0},"minSyncReplicas":{"description":"Minimum number of synchronous replicas required for commit.","type":"integer","default":0}}},"users":{"description":"Users configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"password":{"description":"Password for the user.","type":"string"},"replication":{"description":"Whether the user has replication privileges.","type":"boolean"}}}},"databases":{"description":"Databases configuration map.","type":"object","default":{},"additionalProperties":{"type":"object","properties":{"extensions":{"description":"List of enabled PostgreSQL extensions.","type":"array","items":{"type":"string"}},"roles":{"description":"Roles assigned to users.","type":"object","properties":{"admin":{"description":"List of users with admin privileges.","type":"array","items":{"type":"string"}},"readonly":{"description":"List of users with read-only privileges.","type":"array","items":{"type":"string"}}}}}}},"backup":{"description":"Backup configuration.","type":"object","default":{},"required":["enabled"],"properties":{"destinationPath":{"description":"Destination path for backups (e.g. s3://bucket/path/).","type":"string","default":"s3://bucket/path/to/folder/"},"enabled":{"description":"Enable regular backups.","type":"boolean","default":false},"endpointURL":{"description":"S3 endpoint URL for uploads.","type":"string","default":"http://minio-gateway-service:9000"},"retentionPolicy":{"description":"Retention policy (e.g. \"30d\").","type":"string","default":"30d"},"s3AccessKey":{"description":"Access key for S3 authentication.","type":"string","default":"<your-access-key>"},"s3SecretKey":{"description":"Secret key for S3 authentication.","type":"string","default":"<your-secret-key>"},"schedule":{"description":"Cron schedule for automated backups.","type":"string","default":"0 2 * * * *"}}},"bootstrap":{"description":"Bootstrap configuration.","type":"object","default":{},"required":["enabled","oldName"],"properties":{"enabled":{"description":"Whether to restore from a backup.","type":"boolean","default":false},"oldName":{"description":"Previous cluster name before deletion.","type":"string","default":""},"recoveryTime":{"description":"Timestamp (RFC3339) for point-in-time recovery; empty means latest.","type":"string","default":""},"serverName":{"description":"Barman server name (S3 path prefix) used by the original cluster when writing backups. Set this only when the original cluster had an explicit barmanObjectStore.serverName that differed from its Kubernetes resource name.","type":"string","default":""}}}}}
   release:
     prefix: postgres-
     labels:

From 0fdb25df724ba5d44c794271cc0661d97b1d9b13 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Wed, 22 Apr 2026 11:59:50 +0500
Subject: [PATCH 28/82] ci(api): add codegen drift check

Run root 'make generate' as a pre-commit hook and as a dedicated
CI workflow so missed codegen updates (CRDs, deepcopy, clients, RBAC)
are caught instead of merging stale generated files.

Pre-commit hook is scoped to files that actually affect codegen
(api/, pkg/apis/, hack/update-codegen.sh, hack/boilerplate.go.txt)
so unrelated commits are not slowed down. CI job sets up Go from
go.mod, runs make generate, and fails on drift with a pointer to
the local fix.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/codegen-drift.yml | 42 +++++++++++++++++++++++++++++
 .pre-commit-config.yaml             | 11 ++++++++
 2 files changed, 53 insertions(+)
 create mode 100644 .github/workflows/codegen-drift.yml

diff --git a/.github/workflows/codegen-drift.yml b/.github/workflows/codegen-drift.yml
new file mode 100644
index 00000000..bd7927f7
--- /dev/null
+++ b/.github/workflows/codegen-drift.yml
@@ -0,0 +1,42 @@
+name: Codegen Drift Check
+
+on:
+  pull_request:
+    types: [opened, synchronize, reopened]
+    paths:
+      - 'api/**'
+      - 'pkg/apis/**'
+      - 'hack/update-codegen.sh'
+      - 'hack/boilerplate.go.txt'
+      - 'go.mod'
+      - 'go.sum'
+      - '.github/workflows/codegen-drift.yml'
+
+concurrency:
+  group: codegen-drift-${{ github.workflow }}-${{ github.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  codegen-drift:
+    name: Verify generated code is up to date
+    runs-on: ubuntu-22.04
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+          cache: true
+
+      - name: Run make generate
+        run: make generate
+
+      - name: Fail on drift
+        run: |
+          if ! git diff --exit-code; then
+            echo "::error::'make generate' produced changes. Run 'make generate' locally and commit the result."
+            git status --short
+            exit 1
+          fi
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index ac0f7e30..836f79c5 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -1,6 +1,17 @@
 repos:
 - repo: local
   hooks:
+    - id: run-make-generate-root
+      name: Run 'make generate' at repo root
+      entry: |
+        flock -x .git/pre-commit.lock sh -c '
+          echo "Running make generate at repo root"
+          make generate || exit $?
+          git diff --color=always | cat
+        '
+      language: system
+      files: ^(api/|pkg/apis/|hack/update-codegen\.sh$|hack/boilerplate\.go\.txt$)
+      pass_filenames: false
     - id: run-make-generate
       name: Run 'make generate' in all app directories
       entry: |

From 860f431187c3f577bfd2b7ab60201583bdef9384 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Wed, 22 Apr 2026 12:00:22 +0500
Subject: [PATCH 29/82] chore(api): regenerate deepcopy for
 RestoreJobSpec.Options

The Options field was added to RestoreJobSpec without re-running
'make generate', leaving zz_generated.deepcopy.go out of sync.
Regenerated to include the missing DeepCopyInto handling for the
runtime.RawExtension pointer.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 api/backups/v1alpha1/zz_generated.deepcopy.go | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/api/backups/v1alpha1/zz_generated.deepcopy.go b/api/backups/v1alpha1/zz_generated.deepcopy.go
index 89f6171f..61b8f839 100644
--- a/api/backups/v1alpha1/zz_generated.deepcopy.go
+++ b/api/backups/v1alpha1/zz_generated.deepcopy.go
@@ -620,6 +620,11 @@ func (in *RestoreJobSpec) DeepCopyInto(out *RestoreJobSpec) {
 		*out = new(v1.TypedLocalObjectReference)
 		(*in).DeepCopyInto(*out)
 	}
+	if in.Options != nil {
+		in, out := &in.Options, &out.Options
+		*out = new(runtime.RawExtension)
+		(*in).DeepCopyInto(*out)
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new RestoreJobSpec.

From b52e2801b460e5ba00a864dd5adc29fd4f20fae8 Mon Sep 17 00:00:00 2001
From: "cozystack-ci[bot]"
 <274107086+cozystack-ci[bot]@users.noreply.github.com>
Date: Wed, 22 Apr 2026 07:28:09 +0000
Subject: [PATCH 30/82] Prepare release v1.3.0

Signed-off-by: cozystack-ci[bot] <274107086+cozystack-ci[bot]@users.noreply.github.com>
---
 packages/apps/kubernetes/images/kubevirt-csi-driver.tag     | 2 +-
 packages/apps/mariadb/images/mariadb-backup.tag             | 2 +-
 packages/core/installer/values.yaml                         | 4 ++--
 packages/core/platform/values.yaml                          | 2 +-
 packages/core/testing/values.yaml                           | 2 +-
 packages/extra/bootbox/images/matchbox.tag                  | 2 +-
 packages/extra/seaweedfs/images/objectstorage-sidecar.tag   | 2 +-
 packages/system/backup-controller/values.yaml               | 2 +-
 packages/system/backupstrategy-controller/values.yaml       | 2 +-
 packages/system/bucket/images/s3manager.tag                 | 2 +-
 packages/system/cozystack-api/values.yaml                   | 2 +-
 packages/system/cozystack-controller/values.yaml            | 2 +-
 packages/system/dashboard/templates/configmap.yaml          | 2 +-
 packages/system/dashboard/values.yaml                       | 6 +++---
 .../system/grafana-operator/images/grafana-dashboards.tag   | 2 +-
 packages/system/kamaji/values.yaml                          | 4 ++--
 packages/system/kubeovn-plunger/values.yaml                 | 2 +-
 packages/system/kubeovn-webhook/values.yaml                 | 2 +-
 packages/system/kubevirt-csi-node/values.yaml               | 2 +-
 packages/system/lineage-controller-webhook/values.yaml      | 2 +-
 packages/system/linstor/values.yaml                         | 4 ++--
 packages/system/multus/templates/multus-daemonset-thick.yml | 4 ++--
 packages/system/objectstorage-controller/values.yaml        | 2 +-
 packages/system/seaweedfs/values.yaml                       | 2 +-
 24 files changed, 30 insertions(+), 30 deletions(-)

diff --git a/packages/apps/kubernetes/images/kubevirt-csi-driver.tag b/packages/apps/kubernetes/images/kubevirt-csi-driver.tag
index ea4a21b3..00aa2caa 100644
--- a/packages/apps/kubernetes/images/kubevirt-csi-driver.tag
+++ b/packages/apps/kubernetes/images/kubevirt-csi-driver.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/kubevirt-csi-driver:0.0.0@sha256:36e26a6b9063761ca3f5597bbd6272651d8a502abbc1c4e3a20e6b7e45a2875a
+ghcr.io/cozystack/cozystack/kubevirt-csi-driver:0.0.0@sha256:72154a97054e16cdf3dea6129d962b8d7e86b55cf9386095e8ac2ce7c8b69172
diff --git a/packages/apps/mariadb/images/mariadb-backup.tag b/packages/apps/mariadb/images/mariadb-backup.tag
index 1e381661..6c830892 100644
--- a/packages/apps/mariadb/images/mariadb-backup.tag
+++ b/packages/apps/mariadb/images/mariadb-backup.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/mariadb-backup:0.0.0@sha256:0ddbbec0568dcb9fbc317cd9cc654e826dbe88ba3f184fa9b6b58aacb93b4570
+ghcr.io/cozystack/cozystack/mariadb-backup:0.0.0@sha256:3841eb171416711977dea0cf8cd45d32344caac9727af760c37d5e1dd41ee4bb
diff --git a/packages/core/installer/values.yaml b/packages/core/installer/values.yaml
index f0a363e1..eef691ea 100644
--- a/packages/core/installer/values.yaml
+++ b/packages/core/installer/values.yaml
@@ -1,9 +1,9 @@
 cozystackOperator:
   # Deployment variant: talos, generic, hosted
   variant: talos
-  image: ghcr.io/cozystack/cozystack/cozystack-operator:v1.3.0-rc.1@sha256:d833abf3eac990732440d7d04e2df62ef9cd46704e0637e63e86983132c6d958
+  image: ghcr.io/cozystack/cozystack/cozystack-operator:v1.3.0@sha256:62574f12486bb40c901cf5ed484cca264405ce5810196d86555cbb27cce1ba48
   platformSourceUrl: 'oci://ghcr.io/cozystack/cozystack/cozystack-packages'
-  platformSourceRef: 'digest=sha256:db28afd18635f6295342f58cd90ac7015c8f59d46ae9704fffeb2c654c3c8a0e'
+  platformSourceRef: 'digest=sha256:a0b9ef938446b3132d3d22ad2262beb1027c48c9037b6c2346fdc2f19acd3036'
 # Generic variant configuration (only used when cozystackOperator.variant=generic)
 cozystack:
   # Kubernetes API server host (IP only, no protocol/port)
diff --git a/packages/core/platform/values.yaml b/packages/core/platform/values.yaml
index 1d9a51c5..f33926db 100644
--- a/packages/core/platform/values.yaml
+++ b/packages/core/platform/values.yaml
@@ -5,7 +5,7 @@ sourceRef:
   path: /
 migrations:
   enabled: false
-  image: ghcr.io/cozystack/cozystack/platform-migrations:v1.3.0-rc.1@sha256:555e4b76421361805a84bc9088b01b23a9c4a9430bd8ebd2db82ef9677d7008c
+  image: ghcr.io/cozystack/cozystack/platform-migrations:v1.3.0@sha256:555e4b76421361805a84bc9088b01b23a9c4a9430bd8ebd2db82ef9677d7008c
   targetVersion: 39
 # Bundle deployment configuration
 bundles:
diff --git a/packages/core/testing/values.yaml b/packages/core/testing/values.yaml
index fe76a451..9d4bdd10 100644
--- a/packages/core/testing/values.yaml
+++ b/packages/core/testing/values.yaml
@@ -1,2 +1,2 @@
 e2e:
-  image: ghcr.io/cozystack/cozystack/e2e-sandbox:v1.3.0-rc.1@sha256:0367a03b981df2a3ea13f411d4cb7869c2bf2c89c07d3d5c8971b9a28921ccef
+  image: ghcr.io/cozystack/cozystack/e2e-sandbox:v1.3.0@sha256:0367a03b981df2a3ea13f411d4cb7869c2bf2c89c07d3d5c8971b9a28921ccef
diff --git a/packages/extra/bootbox/images/matchbox.tag b/packages/extra/bootbox/images/matchbox.tag
index 214064c3..f51c7675 100644
--- a/packages/extra/bootbox/images/matchbox.tag
+++ b/packages/extra/bootbox/images/matchbox.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/matchbox:v1.3.0-rc.1@sha256:27959a7e36fb5594049cf88e7b196e661c6ad161dc89330b2c0ef543d1d48367
+ghcr.io/cozystack/cozystack/matchbox:v1.3.0@sha256:85b8e04bf6f0690612dd63e80475df269f4a436d16680f8a40f2860cf16e2f74
diff --git a/packages/extra/seaweedfs/images/objectstorage-sidecar.tag b/packages/extra/seaweedfs/images/objectstorage-sidecar.tag
index 6d85d153..9b413da5 100644
--- a/packages/extra/seaweedfs/images/objectstorage-sidecar.tag
+++ b/packages/extra/seaweedfs/images/objectstorage-sidecar.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/objectstorage-sidecar:v1.3.0-rc.1@sha256:5bcccbdb13979a16cee535eb5fbcdf0d87973689010a10b7b25e55c5be3edaa6
+ghcr.io/cozystack/cozystack/objectstorage-sidecar:v1.3.0@sha256:5bcccbdb13979a16cee535eb5fbcdf0d87973689010a10b7b25e55c5be3edaa6
diff --git a/packages/system/backup-controller/values.yaml b/packages/system/backup-controller/values.yaml
index 3c8834dc..34e033ce 100644
--- a/packages/system/backup-controller/values.yaml
+++ b/packages/system/backup-controller/values.yaml
@@ -1,5 +1,5 @@
 backupController:
-  image: "ghcr.io/cozystack/cozystack/backup-controller:v1.3.0-rc.1@sha256:9d1727b80b436387141b43fd8c12ec36d75220e1d6014ebb58e2e8df3caba5e8"
+  image: "ghcr.io/cozystack/cozystack/backup-controller:v1.3.0@sha256:e1a083dc92f26dfef004f47c1cd20a6357174aad835004f58e751c494b76649a"
   replicas: 2
   debug: false
   metrics:
diff --git a/packages/system/backupstrategy-controller/values.yaml b/packages/system/backupstrategy-controller/values.yaml
index 326dd7c2..d6453124 100644
--- a/packages/system/backupstrategy-controller/values.yaml
+++ b/packages/system/backupstrategy-controller/values.yaml
@@ -1,5 +1,5 @@
 backupStrategyController:
-  image: "ghcr.io/cozystack/cozystack/backupstrategy-controller:v1.3.0-rc.1@sha256:aebdf354e15bff73d4e12d3f6117494002e8a4e586d1c46450cd0a3da493fbb5"
+  image: "ghcr.io/cozystack/cozystack/backupstrategy-controller:v1.3.0@sha256:be0a9ec1f4307064b16388a24628aee46e06252738338add80b99ea1e04e62bf"
   replicas: 2
   debug: false
   metrics:
diff --git a/packages/system/bucket/images/s3manager.tag b/packages/system/bucket/images/s3manager.tag
index 30b082df..889db0af 100644
--- a/packages/system/bucket/images/s3manager.tag
+++ b/packages/system/bucket/images/s3manager.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/s3manager:v0.5.0@sha256:cfffa1267ad72d138c6e9eed6fa282f4fa23698bef2180be98c0be39c073ec28
+ghcr.io/cozystack/cozystack/s3manager:v0.5.0@sha256:fb65e734bbdbdaef2238769cc2ecfb54dbddaf1e0952ad438a9b3e26b9dbb4b5
diff --git a/packages/system/cozystack-api/values.yaml b/packages/system/cozystack-api/values.yaml
index 2ccdb706..dddabf54 100644
--- a/packages/system/cozystack-api/values.yaml
+++ b/packages/system/cozystack-api/values.yaml
@@ -1,3 +1,3 @@
 cozystackAPI:
-  image: ghcr.io/cozystack/cozystack/cozystack-api:v1.3.0-rc.1@sha256:c57f2e62a547e4d62ec1c88708bbce146653bc68e1e312ab16f97a9206bd6750
+  image: ghcr.io/cozystack/cozystack/cozystack-api:v1.3.0@sha256:5fa8648821cf1e9e08cf7c2899c4b1c4226bb74c6773327141456c7d3f2b9b7e
   replicas: 2
diff --git a/packages/system/cozystack-controller/values.yaml b/packages/system/cozystack-controller/values.yaml
index 6addae2b..4d0d0fde 100644
--- a/packages/system/cozystack-controller/values.yaml
+++ b/packages/system/cozystack-controller/values.yaml
@@ -1,4 +1,4 @@
 cozystackController:
-  image: ghcr.io/cozystack/cozystack/cozystack-controller:v1.3.0-rc.1@sha256:5ab50893e9d0237d26f366c9d647da6337ca9b97bae764430571d4fb080f6200
+  image: ghcr.io/cozystack/cozystack/cozystack-controller:v1.3.0@sha256:d03d19b78c4c98f970ac549a68b01ef6bd1ad755f5e0dcb9e08503511cdf2fdc
   debug: false
   disableTelemetry: false
diff --git a/packages/system/dashboard/templates/configmap.yaml b/packages/system/dashboard/templates/configmap.yaml
index c2be8d46..a2b7684c 100644
--- a/packages/system/dashboard/templates/configmap.yaml
+++ b/packages/system/dashboard/templates/configmap.yaml
@@ -1,6 +1,6 @@
 {{- $brandingConfig := .Values._cluster.branding | default dict }}
 
-{{- $tenantText := "v1.3.0-rc.1" }}
+{{- $tenantText := "v1.3.0" }}
 {{- $footerText := "Cozystack" }}
 {{- $titleText := "Cozystack Dashboard" }}
 {{- $logoText := "" }}
diff --git a/packages/system/dashboard/values.yaml b/packages/system/dashboard/values.yaml
index 63305af8..a5188825 100644
--- a/packages/system/dashboard/values.yaml
+++ b/packages/system/dashboard/values.yaml
@@ -1,6 +1,6 @@
 openapiUI:
-  image: ghcr.io/cozystack/cozystack/openapi-ui:v1.3.0-rc.1@sha256:69dcecbea0cfff681043c75e3508c84f03a01b7be3e3a425deb740d629d3ab04
+  image: ghcr.io/cozystack/cozystack/openapi-ui:v1.3.0@sha256:0fa79c373a62840a617ff1ca1b0e31931c13a6cf7b0bb0ff0dc191f047a465a3
 openapiUIK8sBff:
-  image: ghcr.io/cozystack/cozystack/openapi-ui-k8s-bff:v1.3.0-rc.1@sha256:873cf834deddaa3954b6d9c5d520a8e03714b77178eff44826d93540561f599e
+  image: ghcr.io/cozystack/cozystack/openapi-ui-k8s-bff:v1.3.0@sha256:1b3ea6d4c7dbbe6a8def3b2807fffdfab2ac4afc39d7a846e57dd491fa168f92
 tokenProxy:
-  image: ghcr.io/cozystack/cozystack/token-proxy:v1.3.0-rc.1@sha256:2e280991e07853ea48f97b0a42946afffa10d03d6a83d41099ed83e6ffc94fdc
+  image: ghcr.io/cozystack/cozystack/token-proxy:v1.3.0@sha256:2e280991e07853ea48f97b0a42946afffa10d03d6a83d41099ed83e6ffc94fdc
diff --git a/packages/system/grafana-operator/images/grafana-dashboards.tag b/packages/system/grafana-operator/images/grafana-dashboards.tag
index 033a6742..cde83a70 100644
--- a/packages/system/grafana-operator/images/grafana-dashboards.tag
+++ b/packages/system/grafana-operator/images/grafana-dashboards.tag
@@ -1 +1 @@
-ghcr.io/cozystack/cozystack/grafana-dashboards:v1.3.0-rc.1@sha256:2c9aa0b48e2bf6167db198f4d15882bfe51700108edf2e9f6d0942940a2c1204
+ghcr.io/cozystack/cozystack/grafana-dashboards:v1.3.0@sha256:2c9aa0b48e2bf6167db198f4d15882bfe51700108edf2e9f6d0942940a2c1204
diff --git a/packages/system/kamaji/values.yaml b/packages/system/kamaji/values.yaml
index b1ca3cb7..4de74113 100644
--- a/packages/system/kamaji/values.yaml
+++ b/packages/system/kamaji/values.yaml
@@ -3,7 +3,7 @@ kamaji:
     deploy: false
   image:
     pullPolicy: IfNotPresent
-    tag: v1.3.0-rc.1@sha256:8c9af083b60600c0800eb56a2cda75f26007b7272a1cf019140de003bfce1a4d
+    tag: v1.3.0@sha256:8c9af083b60600c0800eb56a2cda75f26007b7272a1cf019140de003bfce1a4d
     repository: ghcr.io/cozystack/cozystack/kamaji
   resources:
     limits:
@@ -13,4 +13,4 @@ kamaji:
       cpu: 100m
       memory: 100Mi
   extraArgs:
-    - --migrate-image=ghcr.io/cozystack/cozystack/kamaji:v1.3.0-rc.1@sha256:8c9af083b60600c0800eb56a2cda75f26007b7272a1cf019140de003bfce1a4d
+    - --migrate-image=ghcr.io/cozystack/cozystack/kamaji:v1.3.0@sha256:8c9af083b60600c0800eb56a2cda75f26007b7272a1cf019140de003bfce1a4d
diff --git a/packages/system/kubeovn-plunger/values.yaml b/packages/system/kubeovn-plunger/values.yaml
index a6946967..02620593 100644
--- a/packages/system/kubeovn-plunger/values.yaml
+++ b/packages/system/kubeovn-plunger/values.yaml
@@ -1,4 +1,4 @@
 portSecurity: true
 routes: ""
-image: ghcr.io/cozystack/cozystack/kubeovn-plunger:v1.3.0-rc.1@sha256:4819095c5237239b8e2282ed1710141a327b3e97773f598f983a22ec0f5add05
+image: ghcr.io/cozystack/cozystack/kubeovn-plunger:v1.3.0@sha256:b75a0facb99c3b0fe8090414b3425f5c4858fff632d1de122ce9944c4daa89c0
 ovnCentralName: ovn-central
diff --git a/packages/system/kubeovn-webhook/values.yaml b/packages/system/kubeovn-webhook/values.yaml
index 24d75d95..0c6ce81a 100644
--- a/packages/system/kubeovn-webhook/values.yaml
+++ b/packages/system/kubeovn-webhook/values.yaml
@@ -1,3 +1,3 @@
 portSecurity: true
 routes: ""
-image: ghcr.io/cozystack/cozystack/kubeovn-webhook:v1.3.0-rc.1@sha256:e18f9fd679e38f65362a8d0042f25468272f6d081136ad47027168d8e7e07a4a
+image: ghcr.io/cozystack/cozystack/kubeovn-webhook:v1.3.0@sha256:e18f9fd679e38f65362a8d0042f25468272f6d081136ad47027168d8e7e07a4a
diff --git a/packages/system/kubevirt-csi-node/values.yaml b/packages/system/kubevirt-csi-node/values.yaml
index a4d51b79..56e28543 100644
--- a/packages/system/kubevirt-csi-node/values.yaml
+++ b/packages/system/kubevirt-csi-node/values.yaml
@@ -1,3 +1,3 @@
 storageClass: replicated
 csiDriver:
-  image: ghcr.io/cozystack/cozystack/kubevirt-csi-driver:0.0.0@sha256:36e26a6b9063761ca3f5597bbd6272651d8a502abbc1c4e3a20e6b7e45a2875a
+  image: ghcr.io/cozystack/cozystack/kubevirt-csi-driver:0.0.0@sha256:72154a97054e16cdf3dea6129d962b8d7e86b55cf9386095e8ac2ce7c8b69172
diff --git a/packages/system/lineage-controller-webhook/values.yaml b/packages/system/lineage-controller-webhook/values.yaml
index 32439941..c5671548 100644
--- a/packages/system/lineage-controller-webhook/values.yaml
+++ b/packages/system/lineage-controller-webhook/values.yaml
@@ -1,5 +1,5 @@
 lineageControllerWebhook:
-  image: ghcr.io/cozystack/cozystack/lineage-controller-webhook:v1.3.0-rc.1@sha256:7e443b9252f0477fe07471f76365edfa28550add51fe8fd4d11312d60a81850e
+  image: ghcr.io/cozystack/cozystack/lineage-controller-webhook:v1.3.0@sha256:e8984709686a5eaf19b89da378d7b8c688ea5607e0783a88d9c9e4ccfca96fb0
   debug: false
   localK8sAPIEndpoint:
     enabled: true
diff --git a/packages/system/linstor/values.yaml b/packages/system/linstor/values.yaml
index 57f5f331..62dd2cac 100644
--- a/packages/system/linstor/values.yaml
+++ b/packages/system/linstor/values.yaml
@@ -1,7 +1,7 @@
 piraeusServer:
   image:
     repository: ghcr.io/cozystack/cozystack/piraeus-server
-    tag: 1.33.2@sha256:e748ea7b2adb285b0387fde4453320155fc03e4470db7d9af1d5447087bf18b4
+    tag: 1.33.2@sha256:553f313ab35dc2e345ef3683156d29e75c23177e2750e9af3a83aa9e23941cbb
 # Talos-specific workarounds (disable for generic Linux like Ubuntu/Debian)
 talos:
   enabled: true
@@ -13,4 +13,4 @@ linstor:
 linstorCSI:
   image:
     repository: ghcr.io/cozystack/cozystack/linstor-csi
-    tag: v1.10.5@sha256:a80d39f96a988085e84be2d6e896f4b4a008fccd5bbd9eaa2b8534b076429942
+    tag: v1.10.5@sha256:b8f59b5659fb1791cb764d3f37df4cf29920aadcc10637231ba7d857233f377d
diff --git a/packages/system/multus/templates/multus-daemonset-thick.yml b/packages/system/multus/templates/multus-daemonset-thick.yml
index 2ddc6362..cad330ea 100644
--- a/packages/system/multus/templates/multus-daemonset-thick.yml
+++ b/packages/system/multus/templates/multus-daemonset-thick.yml
@@ -155,7 +155,7 @@ spec:
       serviceAccountName: multus
       containers:
         - name: kube-multus
-          image: ghcr.io/cozystack/cozystack/multus-cni:v1.3.0-rc.1@sha256:d76cc81c9ef1521e1317aacdc98360325f4fbecdfc687b97852e2f5c275551b1
+          image: ghcr.io/cozystack/cozystack/multus-cni:v1.3.0@sha256:6735ffc12e5e660951f5a42943b38c308f33774a55836e94c191001405b58ec0
           command: [ "/usr/src/multus-cni/bin/multus-daemon" ]
           resources:
             requests:
@@ -201,7 +201,7 @@ spec:
                   fieldPath: spec.nodeName
       initContainers:
         - name: install-multus-binary
-          image: ghcr.io/cozystack/cozystack/multus-cni:v1.3.0-rc.1@sha256:d76cc81c9ef1521e1317aacdc98360325f4fbecdfc687b97852e2f5c275551b1
+          image: ghcr.io/cozystack/cozystack/multus-cni:v1.3.0@sha256:6735ffc12e5e660951f5a42943b38c308f33774a55836e94c191001405b58ec0
           command:
             - "/usr/src/multus-cni/bin/install_multus"
             - "-d"
diff --git a/packages/system/objectstorage-controller/values.yaml b/packages/system/objectstorage-controller/values.yaml
index 4def5b26..a3859194 100644
--- a/packages/system/objectstorage-controller/values.yaml
+++ b/packages/system/objectstorage-controller/values.yaml
@@ -1,3 +1,3 @@
 objectstorage:
   controller:
-    image: "ghcr.io/cozystack/cozystack/objectstorage-controller:v1.3.0-rc.1@sha256:50b7eb360538dec3db747c9b28792b207068defc1f630d4bd965015d934d5279"
+    image: "ghcr.io/cozystack/cozystack/objectstorage-controller:v1.3.0@sha256:7a9e4bf9c3f95f364756396815bb51b6c0f58f85db653460574d94b96cd3c7d5"
diff --git a/packages/system/seaweedfs/values.yaml b/packages/system/seaweedfs/values.yaml
index 57843ff5..3542a77f 100644
--- a/packages/system/seaweedfs/values.yaml
+++ b/packages/system/seaweedfs/values.yaml
@@ -177,7 +177,7 @@ seaweedfs:
     bucketClassName: "seaweedfs"
     region: ""
     sidecar:
-      image: "ghcr.io/cozystack/cozystack/objectstorage-sidecar:v1.3.0-rc.1@sha256:5bcccbdb13979a16cee535eb5fbcdf0d87973689010a10b7b25e55c5be3edaa6"
+      image: "ghcr.io/cozystack/cozystack/objectstorage-sidecar:v1.3.0@sha256:5bcccbdb13979a16cee535eb5fbcdf0d87973689010a10b7b25e55c5be3edaa6"
   certificates:
     commonName: "SeaweedFS CA"
     ipAddresses: []

From 1eeeb2652aaaf6f92e655f010fb3655e865f6abc Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Wed, 22 Apr 2026 15:51:08 +0500
Subject: [PATCH 31/82] docs: add changelog for v1.3.0

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 docs/changelogs/v1.3.0.md | 242 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 242 insertions(+)
 create mode 100644 docs/changelogs/v1.3.0.md

diff --git a/docs/changelogs/v1.3.0.md b/docs/changelogs/v1.3.0.md
new file mode 100644
index 00000000..af7bb30e
--- /dev/null
+++ b/docs/changelogs/v1.3.0.md
@@ -0,0 +1,242 @@
+<!--
+https://github.com/cozystack/cozystack/releases/tag/v1.3.0
+-->
+
+# Cozystack v1.3.0
+
+Cozystack v1.3.0 brings **storage-aware pod scheduling** via a LINSTOR scheduler extender, a managed **LINSTOR GUI** web console with Keycloak SSO, a curated **VM Default Images** catalog for out-of-the-box virtual-machine provisioning, a new **WorkloadsReady / Events** observability surface with S3 bucket metering, and **cross-namespace VMInstance backup restore** with a full **RestoreJob dashboard** flow. The release also ships stricter tenant-name validation, VMInstance network-selector improvements, Keycloak theme injection and SMTP configuration, a host-runtime preflight check, and rolls up every fix from the v1.2.1 → v1.2.4 patch line.
+
+> **Note:** Items marked *(backported to v1.2.x)* were also shipped in v1.2.1, v1.2.2, v1.2.3, or v1.2.4 patch releases.
+
+## Feature Highlights
+
+### Storage-Aware Scheduling via the LINSTOR Extender
+
+The `cozystack-scheduler` now calls a **LINSTOR scheduler extender** for storage-locality-aware pod placement. When a pod declares both a `SchedulingClass` and LINSTOR-backed PVCs, the scheduler consults LINSTOR to prefer nodes where volume replicas already exist — reducing cross-node replication traffic and improving I/O latency for storage-heavy workloads such as databases, object stores, and VMs.
+
+The integration builds on the existing `SchedulingClass` tenant workload placement system introduced in v1.2.0 and requires no tenant-side configuration — workloads simply benefit once a SchedulingClass is assigned. Administrators can mix storage locality with the existing data-center / hardware-generation constraints defined on SchedulingClass CRs ([**@lllamnyp**](https://github.com/lllamnyp) in #2330).
+
+### LINSTOR GUI: Managed Web Console for Storage Administration
+
+A new opt-in `linstor-gui` system package deploys **LINBIT's linstor-gui web UI** alongside the LINSTOR controller with mTLS client authentication, non-root security context, and a ClusterIP-only service by default. When OIDC is configured on the platform, an optional **Keycloak-protected Ingress** (via oauth2-proxy) exposes the UI for browser access. Access is restricted to members of the `cozystack-cluster-admin` Keycloak group, consistent with host-cluster admin RBAC, and the gatekeeper blocks in-app LINSTOR authentication setup at the nginx proxy layer so the managed configuration cannot be subverted through the UI.
+
+Operators who prefer CLI access keep the existing `linstor` command; the GUI is strictly additive and stays disabled by default ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2382, #2390, #2415, #2419).
+
+### VM Default Images: Out-of-the-Box VM Provisioning
+
+The new `vm-default-images` package provides a curated set of **cluster-wide virtual-machine images** (Ubuntu, Debian, CentOS Stream, and others) as pre-populated DataVolumes, so tenants can provision VMs against well-known base images without first having to upload them. The package is opt-in via the `iaas` bundle and defaults to replicated storage for high availability. Migration 38 renames legacy `vm-image-*` DataVolumes to the new `vm-default-images-*` naming scheme, and the `vm-disk` chart gains a new "disk" source type for cloning from existing vm-disks in the same namespace ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2258).
+
+### Application Observability: WorkloadsReady, Events, and S3 Bucket Metering
+
+Applications now expose a **WorkloadsReady** condition on their status by querying associated WorkloadMonitor resources, giving operators a single place to check whether all underlying workloads (Deployments, StatefulSets, DaemonSets, PVCs) are healthy. The dashboard gains a new **Events tab** showing namespace-scoped Kubernetes events per application, with fallback to `.firstTimestamp` when `.eventTime` is absent. A long-standing bug where WorkloadMonitor's `Operational` status was never persisted is fixed in the same change ([**@lexfrei**](https://github.com/lexfrei) in #2356).
+
+The WorkloadMonitor reconciler is extended to track **COSI BucketClaim** objects as first-class Workloads, and the bucket controller now queries SeaweedFS logical and physical bucket-size metrics from VictoriaMetrics via a namespace-scoped monitoring endpoint, enabling S3 billing integration on par with Pods and PVCs ([**@kitsunoff**](https://github.com/kitsunoff) in #2391). Workloads are also enriched with `workloads.cozystack.io/resource-preset` and source-object labels so downstream billing pipelines can correlate monitors with the tenant preset that produced them ([**@androndo**](https://github.com/androndo) in #2416).
+
+### Cross-Namespace VM Backup Restore and RestoreJob Dashboard
+
+The backup system now supports **restoring VMInstance backups into a different namespace** (cross-namespace copy restores) with IP/MAC preservation and safe rename semantics. In-place backup and restore flows for VMDisk and VMInstance are improved: HelmReleases and DataVolumes are properly handled, and Velero failure messages are propagated to the Application status. The backup status structure has been refactored to store underlying resources as a generic opaque JSON object, enabling arbitrary application-specific metadata without status-schema churn ([**@androndo**](https://github.com/androndo) in #2251, #2319, #2329).
+
+The dashboard now ships a complete **RestoreJob experience**: list view, details page, create form, and sidebar entry, with a "Same as backup" fallback rendering when `spec.targetApplicationRef` is omitted. Non-CRD-backed sidebar factories (`kube-*`, `plan`, `backupjob`, `backup`, `restorejob`) are marked static so they pick up consistent managed-by labels across reconciles ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2437).
+
+## Major Features and Improvements
+
+* **[api] Reject tenant names with dashes at Create time**: Enforces alphanumeric-only naming for Tenants at the API level, preventing names with hyphens that would silently fail during Helm reconciliation. A corresponding regex tightening and regression test suite hardens the validation ([**@lexfrei**](https://github.com/lexfrei) in #2380).
+
+* **[platform] Validate computed tenant namespace length**: Rejects Tenant creation when the computed ancestor-chain namespace would exceed the 63-character Kubernetes namespace limit, preventing opaque HelmRelease reconcile errors downstream ([**@lexfrei**](https://github.com/lexfrei) in #2376).
+
+* **[vm-instance] Rename subnets to networks and add dropdown selector**: Renames the misleading `subnets` field to `networks` in VMInstance for clarity, adds a dropdown selector for available networks in the dashboard form, and includes migration 36 to copy existing `subnets` values. The old field remains supported for backward compatibility ([**@sircthulhu**](https://github.com/sircthulhu) in #2263).
+
+* **[keycloak] Enable injecting themes**: Cozystack administrators can now inject custom Keycloak themes via `initContainers` for UI white-labeling and customization ([**@lllamnyp**](https://github.com/lllamnyp) in #2142).
+
+* **[keycloak-configure] Add email verification and SMTP configuration**: Adds configurable Keycloak settings for user self-registration, email verification, and SMTP server configuration, enabling automated user onboarding flows ([**@BROngineer**](https://github.com/BROngineer) in #2318).
+
+* **[postgres] Hardcode PostgreSQL 17 for monitoring databases**: Pins PostgreSQL 17.7 images for system databases (Grafana, Alerta, Harbor, Keycloak, SeaweedFS) and adds migration 37 to backfill `spec.version=v17` for existing PostgreSQL resources, preventing CNPG from defaulting to PostgreSQL 18 *(backported to v1.2.1)* ([**@IvanHunters**](https://github.com/IvanHunters) in #2304).
+
+* **[platform] Prevent installed packages deletion**: Adds the `helm.sh/resource-policy: keep` annotation to platform packages so disabling a package no longer triggers automatic Helm deletion, restoring the documented behavior where operators must explicitly delete a package *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2273).
+
+* **[mariadb] Always enable replication for consistent service naming**: MariaDB now always enables replication, creating `-primary`/`-secondary` services even for single-replica instances. This fixes dashboard visibility and backup functionality for single-replica setups ([**@sircthulhu**](https://github.com/sircthulhu) in #2279).
+
+* **[hack] Add host runtime preflight check**: New `check-host-runtime.sh` script and `make preflight` target that warns operators when a standalone containerd or docker runtime is running alongside the embedded k3s runtime, helping diagnose container-runtime conflicts early in an installation ([**@lexfrei**](https://github.com/lexfrei) in #2371).
+
+* **[hack] Add check-readiness.sh diagnostic script**: A new diagnostic script for tracking platform reconciliation by checking readiness of Packages, ArtifactGenerators, ExternalArtifacts, and HelmReleases, with support for watch mode and continuous monitoring ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2294).
+
+* **[platform] Add resourcePreset labels to WorkloadMonitor labels**: WorkloadMonitor labels with the `workloads.cozystack.io/` prefix are now propagated onto created Workloads; created Workloads always include the reserved `workloads.cozystack.io/monitor` label, and Helm app charts add `workloads.cozystack.io/resource-preset` metadata to WorkloadMonitor manifests, enabling downstream billing pipelines to correlate monitors with the tenant preset that produced them ([**@androndo**](https://github.com/androndo) in #2416).
+
+## Bug Fixes
+
+* **[platform] Migrate ACME HTTP-01 to ingressClassName API**: Switches ACME HTTP-01 issuance from the deprecated `acme.cert-manager.io/http01-ingress-class` annotation to the modern `ingressClassName` field on `ClusterIssuer` and solver pods. Previously, ClusterIssuers referenced a non-existent `nginx` class while each Ingress individually overrode it via annotation — producing `ingressClassName and class cannot be set at the same time` errors when tenants attempted to migrate to the modern field. The migration is atomic: both the ClusterIssuer and consuming Ingresses are updated together *(backported to v1.2.4)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2436).
+
+* **[harbor] Remove incorrect tenant module flags**: Harbor is a PaaS service, not a tenant module. Incorrect `spec.dashboard.module: true` and `internal.cozystack.io/tenantmodule` flags caused Harbor to appear in the sidebar "Modules" section and be misclassified by controllers handling tenant modules. The flags are now removed so Harbor is displayed in its proper PaaS category and is no longer treated as a tenant-scoped HelmRelease ([**@kvaps**](https://github.com/kvaps) in #2444).
+
+* **[kube-ovn] Resolve kubeovn-plunger RBAC forbidden on deployments**: Grants `kube-ovn-plunger` the RBAC needed to list Deployments so it can reconcile `ovn-central`, fixing `deployments.apps is forbidden` errors in `cozy-kubeovn` ([**@kvaps**](https://github.com/kvaps) in #2441).
+
+* **[cilium] Opt-out of cri-containerd.apparmor.d for nsenter init containers**: Opts cilium-agent init containers out of the `cri-containerd.apparmor.d` AppArmor profile on non-Talos variants (`cilium-generic`, `kubeovn-cilium-generic`), fixing `Init:CrashLoopBackOff` on Ubuntu 22.04+ and Debian where the profile denies `nsenter` namespace entry. Talos variants are untouched as Talos does not load the AppArmor LSM *(backported to v1.2.2)* ([**@lexfrei**](https://github.com/lexfrei) in #2370).
+
+* **[virtual-machine] Exclude external VM services from Cilium BPF LB**: Adds the `service.kubernetes.io/service-proxy-name: cozy-proxy` label to VM LoadBalancer services with `external: true`, telling Cilium to skip BPF processing entirely. Fixes inter-tenant connectivity via public LB IPs (Cilium's DNAT caused cross-tenant pod-to-pod flow classification, triggering CiliumClusterwideNetworkPolicy blocks) and restores WholeIP behavior on Cilium 1.19+ where wildcard service drop entries previously blocked traffic to LB IPs on undeclared ports *(backported to v1.2.2)* ([**@mattia-eleuteri**](https://github.com/mattia-eleuteri) in #2357).
+
+* **[monitoring] Fix infra dashboards missing in default variant**: Includes the `cozy-monitoring` namespace in the dashboard rendering condition, fixing infrastructure Grafana dashboards not rendering in the default platform variant (only the `tenant-root` namespace was previously checked) *(backported to v1.2.2)* ([**@mattia-eleuteri**](https://github.com/mattia-eleuteri) in #2365).
+
+* **[postgres] Fix system PostgreSQL images to 17.7-standard-trixie**: Normalizes system PostgreSQL image tags to the `17.7-standard-trixie` variant with migration logic for existing CNPG clusters, ensuring system databases use the correct image variant consistent with the monitoring stack requirements introduced in v1.2.1 *(backported to v1.2.2)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2364).
+
+* **[build] Filter git describe to match only v* tags**: Adds `--match 'v*'` to all `git describe` calls in `hack/common-envs.mk`, preventing the `api/apps/v1alpha1/vX.Y.Z` subtag from being picked up instead of the release tag and producing invalid Docker image tags *(backported to v1.2.2)* ([**@kvaps**](https://github.com/kvaps) in #2386).
+
+* **[platform] Fix resource allocation ratios not propagated to packages**: Restores propagation of `cpuAllocationRatio`, `memoryAllocationRatio`, and `ephemeralStorageAllocationRatio` from `platform/values.yaml` to the `cozystack-values` Secret that managed applications and KubeVirt read, fixing a regression introduced in the bundle restructure that silently ignored operator-configured ratios *(backported to v1.2.1)* ([**@sircthulhu**](https://github.com/sircthulhu) in #2296).
+
+* **[kubernetes] Set explicit ephemeral-storage on virt-launcher pods**: Sets explicit `domain.resources` ephemeral-storage on the VirtualMachine spec to prevent virt-launcher pods from being evicted because LimitRange defaults were too small for the actual emptyDisk capacity *(backported to v1.2.3)* ([**@kvaps**](https://github.com/kvaps) in #2317).
+
+* **[multus] Pin master CNI to 05-cilium.conflist**: Prevents a boot-time race where multus could auto-detect kube-ovn's conflist instead of Cilium's, which would cause pods to bypass the Cilium chain entirely and lose their endpoint *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2315).
+
+* **[multus] Build custom image with DEL cache fix**: Fixes sandbox cleanup deadlock when CNI ADD never completes, preventing stale sandbox name reservations from permanently blocking pod creation after a node disruption *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2313).
+
+* **[linstor] Set verify-alg to crc32c**: Prevents DRBD connection failures on kernels where `crct10dif` is unavailable (e.g., Talos v1.12.6 with kernel 6.18.18) by setting the LINSTOR verify-alg controller default to `crc32c` *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2303).
+
+* **[linstor] Preserve TCP ports during toggle-disk operations**: Saves existing TCP ports into the `LayerPayload` before `removeLayerData()` deletes them, preventing DRBD resources from entering StandAlone state when a satellite misses the resulting update *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2292).
+
+* **[linstor] Increase satellite startup probe failure threshold**: Raises the LINSTOR satellite `startupProbe` `failureThreshold` from 3 to 30 (30s → 300s) in the `LinstorSatelliteConfiguration` pod template, giving satellites with slow storage initialization enough time to come up without being killed and restarted ([**@Arsolitt**](https://github.com/Arsolitt) in #2425).
+
+## Security
+
+* **docs: add SECURITY.md**: Adds vulnerability reporting procedures, disclosure expectations, and supported release lines ([**@kvaps**](https://github.com/kvaps) in #2230).
+
+* **docs: add OpenSSF Best Practices badge to README**: Adds the OpenSSF Best Practices passing badge to the project README ([**@lexfrei**](https://github.com/lexfrei) in #2320).
+
+* **[linstor-gui] Restrict to cozystack-cluster-admin group**: Tightens access control on the `linstor-gui` Ingress so the UI and its underlying LINSTOR controller REST API are reachable only by members of the `cozystack-cluster-admin` Keycloak group. Previously, the oauth2-proxy gatekeeper enforced only realm membership (`--email-domain=*`), allowing any tenant-scoped account to reach the gatekeeper's static mTLS client cert *(backported to release-1.3 via #2419)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2415, #2419).
+
+## Dependencies & Version Updates
+
+* **[kube-ovn] Bump kube-ovn to v1.15.10 with port-group regression fix**: Updates `packages/system/kubeovn` to upstream v1.15.10 (from v1.15.3) and carries a patch for `pkg/controller/pod.go` that preserves a VM LSP's port-group memberships when Kubernetes GCs a completed virt-launcher pod while another virt-launcher pod of the same VM is still running. Without the patch, the destination pod of a successful live migration lost its security groups, network policies, and node-scoped routing until `kube-ovn-controller` was restarted ([**@kvaps**](https://github.com/kvaps) in #2443).
+
+* **[monitoring] Upgrade victoria-metrics-operator to v0.68.4**: Bumps the vendored `victoria-metrics-operator` Helm chart from 0.59.1 to 0.61.0 (operator appVersion v0.68.1 → v0.68.4), picking up upstream fixes for `VMPodScrape` port routing on VMAgent/VLAgent and `StatefulSet` pod deletion (not eviction) when `maxUnavailable=100%` ([**@lexfrei**](https://github.com/lexfrei) in #2426).
+
+* **[linstor] Update piraeus-server to v1.33.2 with selected backports**: Bumps LINSTOR server from v1.33.1 to v1.33.2 with backported patches for stale bitmap adjust retry, LUKS2 header sizing, optimal I/O size detection, and the maintainer implementation. All patches verified against upstream v1.33.2 with `git apply --check` and `gradlew compileJava` *(backported to v1.2.2)* ([**@kvaps**](https://github.com/kvaps) in #2331).
+
+* **[kamaji] Update to 26.3.5-edge, drop upstreamed patches**: Updates Kamaji from edge-26.2.4 to 26.3.5-edge and removes two patches accepted upstream. Adds configurable probe tuning and DataStore readiness conditions ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2260).
+
+* **[talm] Release v0.23.0, v0.23.1, v0.24.0** (github.com/cozystack/talm): Migrates to the Talos v1.12 multi-document machine config format ([**@lexfrei**](https://github.com/lexfrei) in cozystack/talm#116); renders templates online in `apply` to resolve lookups ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/talm#119); bumps dependencies and modernizes the codebase ([**@lexfrei**](https://github.com/lexfrei) in cozystack/talm#124).
+
+* **[ansible-cozystack] Release v1.2.1, v1.2.2, v1.2.4** (github.com/cozystack/ansible-cozystack): Exposes `publishing.externalIPs` and tenant-root ingress via role variables ([**@lexfrei**](https://github.com/lexfrei) in cozystack/ansible-cozystack#30); adds a comprehensive node prerequisites audit ([**@lexfrei**](https://github.com/lexfrei) in cozystack/ansible-cozystack#27); replaces `ansible.utils.ipaddr` with a stdlib-based test plugin ([**@lexfrei**](https://github.com/lexfrei) in cozystack/ansible-cozystack#24); adds `v` prefix to collection version in requirements.yml examples ([**@lexfrei**](https://github.com/lexfrei) in cozystack/ansible-cozystack#23); tracks installer releases v1.2.1 through v1.2.4 ([**@app/renovate**](https://github.com/apps/renovate) in cozystack/ansible-cozystack#20, #22, #29, #31, #32).
+
+## Development, Testing, and CI/CD
+
+* **[ci] Replace cozystack-bot PAT with cozystack-ci GitHub App**: Replaces the long-lived `cozystack-bot` personal access token with short-lived, scoped tokens from the `cozystack-ci` GitHub App across all release workflows (`tags.yaml`, `auto-release.yaml`, `pull-requests-release.yaml`), improving security and auditability of CI operations ([**@tym83**](https://github.com/tym83) in #2351; [**@kvaps**](https://github.com/kvaps) in #2383, #2392).
+
+* **[ci] Add Gemini Code Assist and CodeRabbit configuration**: Adds repository-level configuration for AI code reviewers with ignore patterns for vendored/generated code and incremental review settings ([**@lexfrei**](https://github.com/lexfrei) in #2385).
+
+* **[ci] Promote next/ trunk on new minor/major releases**: Updates `update-website-docs` in `tags.yaml` to match the new docs-versioning contract — the website repo replaces the old "pre-create `vX.Y/` draft directory" scheme with a permanent `content/en/docs/next/` trunk, and released version directories are promoted explicitly by the release workflow ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2433).
+
+* **[tests] Fix Kafka E2E test timeout and retry race condition**: Increases Kafka E2E test timeout from 60s to 300s and fixes a retry race where `kubectl apply` could hit a still-deleting resource ([**@lexfrei**](https://github.com/lexfrei) in #2358).
+
+* **docs: adopt Conventional Commits for commit and PR titles**: Standardizes commit and PR title format to `type(scope): description` across all contributing docs and the PR template ([**@lexfrei**](https://github.com/lexfrei) in #2395).
+
+* **docs(ci): require screenshots for UI changes in PR template**: Adds a mandatory screenshots section to the PR template for UI-related changes ([**@kitsunoff**](https://github.com/kitsunoff) in #2407).
+
+* **chore(maintenance): add @myasnikovdaniil to CODEOWNERS**: Adds @myasnikovdaniil to the default owners in `.github/CODEOWNERS` for automatic review requests ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2434).
+
+## Documentation
+
+* **[website] Add ApplicationDefinition naming convention reference**: Documents how `cozystack-api` resolves kinds to their backing definitions ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#478).
+
+* **[website] Document Talos / talosctl / Cozystack version pairing**: Adds a version compatibility matrix for installation ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#484).
+
+* **[website] Document namespace layout and parent/child derivation**: Explains tenant namespace hierarchy and parent/child namespace derivation rules ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#479).
+
+* **[website] Document the checkbox-then-edit-CR customization pattern for tenants**: Describes the workflow for customizing tenant settings via the CR after initial checkbox-based creation ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#485).
+
+* **[website] Add custom Keycloak themes documentation**: Covers the theme image contract, configuration, `imagePullSecrets`, and theme activation in the Keycloak admin console ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#463).
+
+* **[website] Add bonding (LACP) configuration how-to guide**: Covers network bonding configuration for Cozystack installations ([**@sircthulhu**](https://github.com/sircthulhu) in cozystack/website#459).
+
+* **[website] Improve registry mirrors for tenant Kubernetes in air-gapped guide**: Improves documentation for configuring registry mirrors in air-gapped environments ([**@sircthulhu**](https://github.com/sircthulhu) in cozystack/website#461).
+
+* **[website] Rewrite guide for ApplicationDefinition API (external-apps)**: Comprehensive rewrite of the external apps guide using the ApplicationDefinition API with Minecraft server examples ([**@kitsunoff**](https://github.com/kitsunoff) in cozystack/website#488).
+
+* **[website] Add documentation for Go types usage**: Guide for using generated Go types for Cozystack managed applications as a Go module ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#465).
+
+* **[website] Update backup/restore documentation for VMI/VMDisk**: Updates backup documentation with VM instance and VM disk restore improvements ([**@androndo**](https://github.com/androndo) in cozystack/website#466).
+
+* **[website] Refactor docs versions to major.minor variants**: Moves docs to major.minor versioning for the v1.x series ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#477).
+
+* **[website] Trunk-based versioning with permanent next/ directory**: Replaces the old "pre-create `vX.Y/` draft directory" scheme with a permanent `content/en/docs/next/` trunk; released version directories are promoted explicitly by `hack/release_next.sh` on new minor/major releases, and routing between `next/` and `vX.Y/` is Makefile-driven ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#495).
+
+* **[website] Add updated OpenAPI spec**: Updates the OpenAPI specification for managed applications reference ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#469).
+
+* **[website] Add OpenAPI spec download to GitHub Pages build**: Fixes the GitHub Pages build to include the OpenAPI spec download ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#494).
+
+* **[website] Add OSS Health pages and OpenSSF badge**: Adds OSS Health section with OpenSSF Scorecard and Best Practices badges to the website ([**@tym83**](https://github.com/tym83) in cozystack/website#470).
+
+* **[website] Add Telemetry page under OSS Health section**: Adds the Telemetry page with initial data seeding to the OSS Health docs ([**@tym83**](https://github.com/tym83) in cozystack/website#471, cozystack/website#504).
+
+* **[website] Blog: OSS Health section launch announcement**: Publishes the announcement blog post for the OSS Health section ([**@tym83**](https://github.com/tym83) in cozystack/website#474).
+
+* **[website] Fix OpenSSF canonical status URL**: Changes the OpenSSF canonical status URL from pt-BR to en ([**@tym83**](https://github.com/tym83) in cozystack/website#475).
+
+* **[website] Add CozySummit Virtual 2026 program announcement**: Publishes the CozySummit Virtual 2026 program announcement blog post ([**@tym83**](https://github.com/tym83) in cozystack/website#472).
+
+* **[website] Add missing release announcements for v0.1–v0.41**: Backfills missing release announcement blog posts for historical Cozystack versions ([**@tym83**](https://github.com/tym83) in cozystack/website#468).
+
+* **[website] Blog: managed PostgreSQL with synchronous replication**: Adds a post covering the managed PostgreSQL synchronous-replication feature ([**@tym83**](https://github.com/tym83) in cozystack/website#497).
+
+* **[website] Blog taxonomies and client-side filter UI**: Registers article-type and topic taxonomies and adds a client-side filter on the blog list page ([**@tym83**](https://github.com/tym83) in cozystack/website#499).
+
+* **[website] Add images frontmatter for social preview on existing posts**: Adds images frontmatter for social preview on existing blog posts ([**@tym83**](https://github.com/tym83) in cozystack/website#498).
+
+* **[website] Fix broken links and stale anchors across v1 docs**: Fixes 14 broken links and stale talm anchors ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#486).
+
+* **[website] Prefix bundle package names with cozystack. in v1 examples**: Corrects package naming in documentation examples ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#482).
+
+* **[website] Finish isolated-field removal and document opt-in policy labels**: Removes the obsolete `isolated` field from tenant documentation and documents the new opt-in policy labels approach ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#481).
+
+* **[website] Add --take-ownership flag and describe networking.* fields**: Documents the `--take-ownership` flag and `networking.*` fields in the installation guide ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#480).
+
+* **[website] Fix KubeOVN MASTER_NODES example path and key in troubleshooting**: Corrects the MASTER_NODES example path and key ([**@lexfrei**](https://github.com/lexfrei) in cozystack/website#483).
+
+* **[website] Add CLAUDE.md for AI agent guidance**: Adds a CLAUDE.md file describing the trunk-based docs architecture for AI agent guidance ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#489).
+
+* **[website] Update /docs/v1/ redirect to latest v1.2**: Updates the `/docs/v1/` redirect target to point to the latest v1.2 docs on GitHub Pages ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#492).
+
+* **[website] Remove nbykov from CODEOWNERS and CLAUDE.md**: Cleans up CODEOWNERS and CLAUDE.md entries ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in cozystack/website#491).
+
+* **[website] Add Ahrefs Analytics tracker**: Adds the Ahrefs Analytics tracker to the website ([**@tym83**](https://github.com/tym83) in cozystack/website#503).
+
+* **[website] Add breathing room between navbar and hero on OSS Health**: Minor styling fix for the OSS Health section ([**@tym83**](https://github.com/tym83) in cozystack/website#500).
+
+* **[website] Fix og social badge image and title**: Updates the social badge image and title ([**@tym83**](https://github.com/tym83) in cozystack/website#487).
+
+* **[website] Update managed apps reference for v1.2.1**: Automated managed-apps reference update ([**@cozystack-bot**](https://github.com/cozystack-bot) in cozystack/website#464).
+
+* **[external-apps-example] Replace MongoDB example with Minecraft apps**: Refactors the external apps example to use the ApplicationDefinition API with Minecraft server applications ([**@lexfrei**](https://github.com/lexfrei) in cozystack/external-apps-example#2).
+
+* **docs: update README introductory description**: Refines the platform positioning and improves clarity on core capabilities in the main README ([**@tym83**](https://github.com/tym83) in #2409).
+
+## Governance
+
+* **Add Mattia Eleuteri ([@mattia-eleuteri](https://github.com/mattia-eleuteri)) as Maintainer**: CSI, Storage, Networking & Security ([**@tym83**](https://github.com/tym83) in #2345).
+
+* **Add Matthieu Robin ([@matthieu-robin](https://github.com/matthieu-robin)) as Maintainer**: Managed applications, platform quality, and benchmarking ([**@tym83**](https://github.com/tym83) in #2346).
+
+## Contributors
+
+We'd like to thank all contributors who made this release possible:
+
+* [**@androndo**](https://github.com/androndo)
+* [**@Arsolitt**](https://github.com/Arsolitt)
+* [**@BROngineer**](https://github.com/BROngineer)
+* [**@IvanHunters**](https://github.com/IvanHunters)
+* [**@kitsunoff**](https://github.com/kitsunoff)
+* [**@kvaps**](https://github.com/kvaps)
+* [**@lexfrei**](https://github.com/lexfrei)
+* [**@lllamnyp**](https://github.com/lllamnyp)
+* [**@mattia-eleuteri**](https://github.com/mattia-eleuteri)
+* [**@myasnikovdaniil**](https://github.com/myasnikovdaniil)
+* [**@sircthulhu**](https://github.com/sircthulhu)
+* [**@tym83**](https://github.com/tym83)
+
+### New Contributors
+
+We're excited to welcome our first-time contributors:
+
+* [**@Arsolitt**](https://github.com/Arsolitt) — First contribution!
+
+---
+
+**Full Changelog**: https://github.com/cozystack/cozystack/compare/v1.2.0...v1.3.0

From 44bc79cef115c02e0e586955c29ee184557c52f3 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 10:17:30 +0500
Subject: [PATCH 32/82] docs(changelog): correct v1.3.0 postgres and
 linstor-gui entries

Post-release cleanup of docs/changelogs/v1.3.0.md so the notes match
what users actually experience in the released v1.3.0:

- Rewrite the postgres major-features entry so author (myasnikovdaniil),
  PR (#2369), and description all match the 17.7-standard-trixie pin +
  migration-37 imageName rewrite that actually shipped. The previous
  entry credited #2304 (superseded spec.version=v17 backfill approach).
- Remove the duplicate #2364 postgres bug-fix entry; the same work is
  now folded into the single major-features entry above, with backport
  references to #2309 (v1.2.1) and #2364 (v1.2.2).
- Remove the [linstor-gui] Restrict to cozystack-cluster-admin group
  security entry. The vulnerable state never shipped in a tagged
  release, so there is nothing user-facing to announce; the restriction
  is already described in the linstor-gui Feature Highlights section
  as part of the feature's day-one behavior.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 docs/changelogs/v1.3.0.md | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/docs/changelogs/v1.3.0.md b/docs/changelogs/v1.3.0.md
index af7bb30e..60c800a8 100644
--- a/docs/changelogs/v1.3.0.md
+++ b/docs/changelogs/v1.3.0.md
@@ -50,7 +50,7 @@ The dashboard now ships a complete **RestoreJob experience**: list view, details
 
 * **[keycloak-configure] Add email verification and SMTP configuration**: Adds configurable Keycloak settings for user self-registration, email verification, and SMTP server configuration, enabling automated user onboarding flows ([**@BROngineer**](https://github.com/BROngineer) in #2318).
 
-* **[postgres] Hardcode PostgreSQL 17 for monitoring databases**: Pins PostgreSQL 17.7 images for system databases (Grafana, Alerta, Harbor, Keycloak, SeaweedFS) and adds migration 37 to backfill `spec.version=v17` for existing PostgreSQL resources, preventing CNPG from defaulting to PostgreSQL 18 *(backported to v1.2.1)* ([**@IvanHunters**](https://github.com/IvanHunters) in #2304).
+* **[postgres] Pin system PostgreSQL to 17.7-standard-trixie**: Pins the PostgreSQL image for system databases (Grafana, Alerta, Harbor, Keycloak, SeaweedFS) to `17.7-standard-trixie` across chart templates and `values.yaml`, and ships migration 37 to patch existing CNPG Cluster `imageName` fields to the same variant (handling unset, any PG 17 tag, and bare-version tags). This prevents CNPG from defaulting to PostgreSQL 18 and locks system databases to the trixie variant consistent with the monitoring stack requirements *(related backports shipped in v1.2.1 via #2309 and v1.2.2 via #2364)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2369).
 
 * **[platform] Prevent installed packages deletion**: Adds the `helm.sh/resource-policy: keep` annotation to platform packages so disabling a package no longer triggers automatic Helm deletion, restoring the documented behavior where operators must explicitly delete a package *(backported to v1.2.1)* ([**@kvaps**](https://github.com/kvaps) in #2273).
 
@@ -76,8 +76,6 @@ The dashboard now ships a complete **RestoreJob experience**: list view, details
 
 * **[monitoring] Fix infra dashboards missing in default variant**: Includes the `cozy-monitoring` namespace in the dashboard rendering condition, fixing infrastructure Grafana dashboards not rendering in the default platform variant (only the `tenant-root` namespace was previously checked) *(backported to v1.2.2)* ([**@mattia-eleuteri**](https://github.com/mattia-eleuteri) in #2365).
 
-* **[postgres] Fix system PostgreSQL images to 17.7-standard-trixie**: Normalizes system PostgreSQL image tags to the `17.7-standard-trixie` variant with migration logic for existing CNPG clusters, ensuring system databases use the correct image variant consistent with the monitoring stack requirements introduced in v1.2.1 *(backported to v1.2.2)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2364).
-
 * **[build] Filter git describe to match only v* tags**: Adds `--match 'v*'` to all `git describe` calls in `hack/common-envs.mk`, preventing the `api/apps/v1alpha1/vX.Y.Z` subtag from being picked up instead of the release tag and producing invalid Docker image tags *(backported to v1.2.2)* ([**@kvaps**](https://github.com/kvaps) in #2386).
 
 * **[platform] Fix resource allocation ratios not propagated to packages**: Restores propagation of `cpuAllocationRatio`, `memoryAllocationRatio`, and `ephemeralStorageAllocationRatio` from `platform/values.yaml` to the `cozystack-values` Secret that managed applications and KubeVirt read, fixing a regression introduced in the bundle restructure that silently ignored operator-configured ratios *(backported to v1.2.1)* ([**@sircthulhu**](https://github.com/sircthulhu) in #2296).
@@ -100,8 +98,6 @@ The dashboard now ships a complete **RestoreJob experience**: list view, details
 
 * **docs: add OpenSSF Best Practices badge to README**: Adds the OpenSSF Best Practices passing badge to the project README ([**@lexfrei**](https://github.com/lexfrei) in #2320).
 
-* **[linstor-gui] Restrict to cozystack-cluster-admin group**: Tightens access control on the `linstor-gui` Ingress so the UI and its underlying LINSTOR controller REST API are reachable only by members of the `cozystack-cluster-admin` Keycloak group. Previously, the oauth2-proxy gatekeeper enforced only realm membership (`--email-domain=*`), allowing any tenant-scoped account to reach the gatekeeper's static mTLS client cert *(backported to release-1.3 via #2419)* ([**@myasnikovdaniil**](https://github.com/myasnikovdaniil) in #2415, #2419).
-
 ## Dependencies & Version Updates
 
 * **[kube-ovn] Bump kube-ovn to v1.15.10 with port-group regression fix**: Updates `packages/system/kubeovn` to upstream v1.15.10 (from v1.15.3) and carries a patch for `pkg/controller/pod.go` that preserves a VM LSP's port-group memberships when Kubernetes GCs a completed virt-launcher pod while another virt-launcher pod of the same VM is still running. Without the patch, the destination pod of a successful live migration lost its security groups, network policies, and node-scoped routing until `kube-ovn-controller` was restarted ([**@kvaps**](https://github.com/kvaps) in #2443).

From e1c6f9c0299c36aae80e71ea7ca9da0eef050a58 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 10:22:09 +0500
Subject: [PATCH 33/82] docs(agents): scope changelog.md to a file-only
 deliverable
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The v1.3.0 release pipeline broke because Copilot, invoked by
.github/workflows/tags.yaml with --allow-all-tools, committed the
generated changelog onto HEAD of main on its own. The workflow's
next step — `git checkout -b ... origin/main` — then wiped the file,
and `git add` failed with a pathspec error.

The root cause is in this document. The checklist ends with "Save
the changelog", which an agent with broad tool access can reasonably
interpret as "also commit it, push it, and open a PR". There was no
explicit boundary.

Add a "Scope and boundaries" section at the top and an explicit
"then exit" at the end of Step 9:

- The single deliverable is docs/changelogs/v<version>.md.
- Forbidden by default: git commit / push / checkout (to switch
  branches) / branch / tag / reset / merge / rebase; PR creation;
  GitHub API writes (POST/PATCH/DELETE); modifying any file other
  than the changelog.
- Read-only analysis (git log/show/fetch/diff, gh pr view, gh api
  GET) remains expected.
- Auxiliary repo clones under _repos/ remain allowed for cross-repo
  analysis per Step 6.
- Scoped "unless the caller explicitly instructs otherwise" so
  interactive use with an IDE remains flexible.

With the rules in the doc, CI and interactive callers share the
same boundary; the workflow can invoke the doc with a one-line
prompt instead of re-stating the constraints every time.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 docs/agents/changelog.md | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/docs/agents/changelog.md b/docs/agents/changelog.md
index 35d59a00..13a674a2 100644
--- a/docs/agents/changelog.md
+++ b/docs/agents/changelog.md
@@ -6,6 +6,20 @@ This file contains detailed instructions for AI-powered IDE on how to generate c
 
 Follow these instructions when the user explicitly asks to generate a changelog.
 
+## Scope and boundaries
+
+**Your single deliverable is the file `docs/changelogs/v<version>.md`.** Write the complete, verified changelog to that path. That is the entire task. Exit as soon as the file is written and verified against the checklist in Step 9.
+
+Unless the caller explicitly instructs otherwise:
+
+- **Do not** run `git commit`, `git push`, `git checkout` (to switch branches), `git branch`, `git tag`, `git reset`, `git merge`, `git rebase`, or any other command that writes to refs, HEAD, or remotes.
+- **Do not** create pull requests, push branches, or issue GitHub API write calls (POST / PATCH / DELETE).
+- In the cozystack working tree, the **only** file you create or modify is `docs/changelogs/v<version>.md`. Cloning auxiliary repositories under `_repos/` for cross-repo analysis (see Step 6) is fine — that directory is outside the cozystack tree.
+
+The caller — a GitHub Actions workflow in CI, or a developer running you interactively — owns branching, committing, pushing, and PR creation. They will perform those actions after you exit. Do not pre-empt them even if the working tree looks ready.
+
+Read-only analysis is expected and encouraged: `git log`, `git show`, `git fetch`, `git diff`, `gh pr view`, `gh api` GET requests, and reading any file in the repository.
+
 ## Required Tools
 
 Before generating changelogs, ensure you have access to `gh` (GitHub CLI) tool, which is used to fetch commit and PR author information. The GitHub CLI is used to correctly identify PR authors from commits and pull requests.
@@ -608,6 +622,8 @@ Create a new changelog file in the format matching previous versions:
 **Save the changelog:**
 Save the changelog to file `docs/changelogs/v<version>.md` according to the version for which the changelog is being generated.
 
+**Then exit.** Do not commit, push, create a branch, or open a pull request — the caller handles all git and GitHub operations after you return. See the "Scope and boundaries" section at the top of this document.
+
 ### Important notes
 
 - **After fetch with --force** local tags are up-to-date, use them for work

From 3720f0f3f2306db9faa4bf1a68bf928e319a04a3 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 10:22:26 +0500
Subject: [PATCH 34/82] fix(ci): harden tags.yaml changelog job against agent
 misbehavior

Three changes to the generate-changelog job to fix the v1.3.0
release pipeline failure (run 24765377017) and make the job robust
to whatever state the Copilot step leaves behind.

1. Add `timeout-minutes: 30` to the Generate changelog using AI
   step. On the v1.3.0 re-run the step hung silently for 10+
   minutes; with no timeout a hung Copilot would hold a self-hosted
   runner for up to 6 hours (job default). The previous successful
   run took ~26 minutes, so 30 is a reasonable ceiling.

2. Replace the terse, ambiguous Copilot prompt with a one-liner
   that invokes docs/agents/changelog.md directly. The "Scope and
   boundaries" section added to that doc in the previous commit is
   now the single source of truth for what the agent may and may
   not do, so the workflow only needs to pass the version and
   point at the relevant doc. VERSION is moved to step env: to
   match GitHub's workflow-injection hardening guidance.

3. Rewrite the Create changelog branch and commit step:
   - add `set -euo pipefail` so any failure is visible
   - validate the file exists up front and fail loud with
     `::error::` if not
   - copy the file to a tempfile BEFORE `git checkout -b`, so the
     checkout to `origin/main` cannot remove it (this is the fix
     for the original pathspec error the v1.3.0 run hit)
   - use `trap` to clean up the tempfile on any exit path
   - move VERSION to env
   - drop the dead "no changes to commit" branch: the check_changelog
     step earlier in the job gates this step on the file being
     absent from origin/main, so `git add` + `git commit` must
     produce a diff. If they don't (e.g. Copilot emitted an empty
     file), fail loud instead of pushing an empty branch.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/tags.yaml | 67 ++++++++++++++++++++-----------------
 1 file changed, 36 insertions(+), 31 deletions(-)

diff --git a/.github/workflows/tags.yaml b/.github/workflows/tags.yaml
index b10897cf..0ec0217d 100644
--- a/.github/workflows/tags.yaml
+++ b/.github/workflows/tags.yaml
@@ -303,51 +303,56 @@ jobs:
 
       - name: Generate changelog using AI
         if: steps.check_changelog.outputs.exists == 'false'
+        timeout-minutes: 30
         env:
           COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_GITHUB_TOKEN }}
           GH_TOKEN: ${{ steps.app-token.outputs.token }}
+          VERSION: ${{ steps.tag.outputs.version }}
         run: |
-          copilot --prompt "prepare changelog file for tagged release v${{ steps.tag.outputs.version }}, use @docs/agents/changelog.md for it. Create the changelog file at docs/changelogs/v${{ steps.tag.outputs.version }}.md" \
+          copilot \
+            --prompt "Generate the release changelog for tag v${VERSION}. Follow the instructions in @docs/agents/changelog.md exactly, including the 'Scope and boundaries' section at the top. Your deliverable is the single file docs/changelogs/v${VERSION}.md — write it and exit; this workflow handles branching, committing, pushing, and opening the PR." \
             --allow-all-tools --allow-all-paths < /dev/null
 
       - name: Create changelog branch and commit
         if: steps.check_changelog.outputs.exists == 'false'
         env:
           APP_TOKEN: ${{ steps.app-token.outputs.token }}
+          VERSION: ${{ steps.tag.outputs.version }}
         run: |
-          git config user.name  "cozystack-ci[bot]"
-          git config user.email "274107086+cozystack-ci[bot]@users.noreply.github.com"
-          git remote set-url origin https://x-access-token:${APP_TOKEN}@github.com/${GITHUB_REPOSITORY}
-          
-          CHANGELOG_FILE="docs/changelogs/v${{ steps.tag.outputs.version }}.md"
-          CHANGELOG_BRANCH="changelog-v${{ steps.tag.outputs.version }}"
-          
-          if [ -f "$CHANGELOG_FILE" ]; then
-            # Fetch latest main branch
-            git fetch origin main
-            
-            # Delete local branch if it exists
-            git branch -D "$CHANGELOG_BRANCH" 2>/dev/null || true
-            
-            # Create and checkout new branch from main
-            git checkout -b "$CHANGELOG_BRANCH" origin/main
-            
-            # Add and commit changelog
-            git add "$CHANGELOG_FILE"
-            if git diff --staged --quiet; then
-              echo "⚠️ No changes to commit (file may already be committed)"
-            else
-              git commit -m "docs: add changelog for v${{ steps.tag.outputs.version }}" -s
-              echo "✅ Changelog committed to branch $CHANGELOG_BRANCH"
-            fi
-            
-            # Push the branch (force push to update if it exists)
-            git push -f origin "$CHANGELOG_BRANCH"
-          else
-            echo "⚠️ Changelog file was not generated"
+          set -euo pipefail
+
+          CHANGELOG_FILE="docs/changelogs/v${VERSION}.md"
+          CHANGELOG_BRANCH="changelog-v${VERSION}"
+
+          if [ ! -f "$CHANGELOG_FILE" ]; then
+            echo "::error::Changelog file $CHANGELOG_FILE was not produced by the Generate changelog using AI step"
             exit 1
           fi
 
+          # Snapshot the file across the branch switch — the checkout below
+          # resets tracked files to match origin/main.
+          TEMP_FILE="$(mktemp)"
+          trap 'rm -f "$TEMP_FILE"' EXIT
+          cp "$CHANGELOG_FILE" "$TEMP_FILE"
+
+          git config user.name  "cozystack-ci[bot]"
+          git config user.email "274107086+cozystack-ci[bot]@users.noreply.github.com"
+          git remote set-url origin "https://x-access-token:${APP_TOKEN}@github.com/${GITHUB_REPOSITORY}"
+
+          git fetch origin main
+          git branch -D "$CHANGELOG_BRANCH" 2>/dev/null || true
+          git checkout -b "$CHANGELOG_BRANCH" origin/main
+
+          mkdir -p "$(dirname "$CHANGELOG_FILE")"
+          cp "$TEMP_FILE" "$CHANGELOG_FILE"
+
+          # The `check_changelog` step gated this job on the file being absent
+          # from origin/main, so `git add` + `git commit` must produce a diff.
+          # If they don't, something is wrong (e.g. empty file) — fail loud.
+          git add "$CHANGELOG_FILE"
+          git commit -m "docs: add changelog for v${VERSION}" -s
+          git push -f origin "$CHANGELOG_BRANCH"
+
       - name: Create PR for changelog
         if: steps.check_changelog.outputs.exists == 'false'
         uses: actions/github-script@v7

From c4477259c7e914accaf45238a2b2b591451672a9 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 10:41:41 +0500
Subject: [PATCH 35/82] fix(backups): move velero-configmap Role to velero
 chart

The backupstrategy-controller chart declared a Role/RoleBinding in the
cozy-velero namespace for ResourceModifier ConfigMap management. Because
velero is an optional package, that namespace does not exist in bundles
without velero, so Helm install aborted with "namespaces \"cozy-velero\"
not found" and blocked the default install of backupstrategy-controller.

Move the Role and RoleBinding into the velero chart so they are created
only when velero is actually installed. The RoleBinding subject points
to the backupstrategy-controller ServiceAccount in its fixed namespace
(cozy-backup-controller).

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .../templates/rbac-bind.yaml                  | 14 ----------
 .../templates/rbac.yaml                       | 15 +++--------
 .../backupstrategy-controller-rbac.yaml       | 26 +++++++++++++++++++
 3 files changed, 29 insertions(+), 26 deletions(-)
 create mode 100644 packages/system/velero/templates/backupstrategy-controller-rbac.yaml

diff --git a/packages/system/backupstrategy-controller/templates/rbac-bind.yaml b/packages/system/backupstrategy-controller/templates/rbac-bind.yaml
index 4ecacce0..03578cc2 100644
--- a/packages/system/backupstrategy-controller/templates/rbac-bind.yaml
+++ b/packages/system/backupstrategy-controller/templates/rbac-bind.yaml
@@ -10,17 +10,3 @@ subjects:
 - kind: ServiceAccount
   name: backupstrategy-controller
   namespace: {{ .Release.Namespace }}
----
-apiVersion: rbac.authorization.k8s.io/v1
-kind: RoleBinding
-metadata:
-  name: backups.cozystack.io:strategy-controller:velero-configmaps
-  namespace: cozy-velero
-roleRef:
-  apiGroup: rbac.authorization.k8s.io
-  kind: Role
-  name: backups.cozystack.io:strategy-controller:velero-configmaps
-subjects:
-- kind: ServiceAccount
-  name: backupstrategy-controller
-  namespace: {{ .Release.Namespace }}
diff --git a/packages/system/backupstrategy-controller/templates/rbac.yaml b/packages/system/backupstrategy-controller/templates/rbac.yaml
index bbdb74f8..634ea88b 100644
--- a/packages/system/backupstrategy-controller/templates/rbac.yaml
+++ b/packages/system/backupstrategy-controller/templates/rbac.yaml
@@ -27,7 +27,9 @@ rules:
   resources: ["pods"]
   verbs: ["get", "list", "watch"]
 # ConfigMaps: controller-runtime cache requires cluster-scoped list/watch;
-# create/update/delete is scoped to cozy-velero via the Role below.
+# create/update/delete is scoped to cozy-velero via a Role shipped by the
+# velero chart (packages/system/velero/templates/backupstrategy-controller-rbac.yaml)
+# so it is only created when velero is installed.
 - apiGroups: [""]
   resources: ["configmaps"]
   verbs: ["get", "list", "watch"]
@@ -78,14 +80,3 @@ rules:
 - apiGroups: ["coordination.k8s.io"]
   resources: ["leases"]
   verbs: ["get", "list", "watch", "create", "update", "patch"]
----
-# To create ResourceModifiers in ConfigMaps for Restore in Velero install namespace.
-apiVersion: rbac.authorization.k8s.io/v1
-kind: Role
-metadata:
-  name: backups.cozystack.io:strategy-controller:velero-configmaps
-  namespace: cozy-velero
-rules:
-- apiGroups: [""]
-  resources: ["configmaps"]
-  verbs: ["create", "get", "list", "delete", "deletecollection", "patch", "update"]
diff --git a/packages/system/velero/templates/backupstrategy-controller-rbac.yaml b/packages/system/velero/templates/backupstrategy-controller-rbac.yaml
new file mode 100644
index 00000000..43c51cd7
--- /dev/null
+++ b/packages/system/velero/templates/backupstrategy-controller-rbac.yaml
@@ -0,0 +1,26 @@
+# Grants the backupstrategy-controller permission to manage ResourceModifier
+# ConfigMaps in the Velero install namespace. Lives here (not in the
+# backupstrategy-controller chart) so that the Role is only created when
+# velero is actually installed — otherwise the chart would try to create it
+# in a namespace that does not exist.
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: backups.cozystack.io:strategy-controller:velero-configmaps
+rules:
+- apiGroups: [""]
+  resources: ["configmaps"]
+  verbs: ["create", "get", "list", "delete", "deletecollection", "patch", "update"]
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: backups.cozystack.io:strategy-controller:velero-configmaps
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: backups.cozystack.io:strategy-controller:velero-configmaps
+subjects:
+- kind: ServiceAccount
+  name: backupstrategy-controller
+  namespace: cozy-backup-controller

From 3c95f3052161372cc0c53c4b7119b56118d19846 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 12:25:19 +0500
Subject: [PATCH 36/82] docs(agents): scope git-write ban to cozystack working
 tree
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Review feedback on PR #2460: the previous "do not write to refs, HEAD,
or remotes" wording contradicted the explicit allowance of `git fetch`
(which updates remote-tracking refs) and the mandatory cross-repo
checks in Step 6, which `cd` into `_repos/<repo>` and run
`git checkout`, `git pull`, etc.

Sharpen the scope paragraph:

- The git-write ban is now explicitly scoped to the cozystack working
  tree — it bans writing to local branches, tags, or HEAD in that
  repo, not "refs/HEAD/remotes" globally.
- `git fetch` is called out as expected.
- Local git operations inside disposable `_repos/` clones
  (`git checkout`, `git pull`, etc.) are explicitly allowed, with
  the remaining rules (no push, no PR creation, no API writes)
  applying to any repository.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 docs/agents/changelog.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/docs/agents/changelog.md b/docs/agents/changelog.md
index 13a674a2..2fde1224 100644
--- a/docs/agents/changelog.md
+++ b/docs/agents/changelog.md
@@ -12,9 +12,9 @@ Follow these instructions when the user explicitly asks to generate a changelog.
 
 Unless the caller explicitly instructs otherwise:
 
-- **Do not** run `git commit`, `git push`, `git checkout` (to switch branches), `git branch`, `git tag`, `git reset`, `git merge`, `git rebase`, or any other command that writes to refs, HEAD, or remotes.
-- **Do not** create pull requests, push branches, or issue GitHub API write calls (POST / PATCH / DELETE).
-- In the cozystack working tree, the **only** file you create or modify is `docs/changelogs/v<version>.md`. Cloning auxiliary repositories under `_repos/` for cross-repo analysis (see Step 6) is fine — that directory is outside the cozystack tree.
+- **In the cozystack working tree**, do not run `git commit`, `git push`, `git checkout` (to switch branches), `git branch`, `git tag`, `git reset`, `git merge`, or `git rebase`. Do not write to local branches, tags, or HEAD. `git fetch` is expected and fine (see the read-only analysis list below).
+- **Do not** push to any remote, open pull requests, or issue GitHub API write calls (POST / PATCH / DELETE) for any repository.
+- In the cozystack working tree, the **only** file you create or modify is `docs/changelogs/v<version>.md`. Cloning auxiliary repositories under `_repos/` for cross-repo analysis (see Step 6) is fine; local git operations inside those disposable clones (`git checkout`, `git pull`, etc.) are allowed — just never push from them or open PRs against them.
 
 The caller — a GitHub Actions workflow in CI, or a developer running you interactively — owns branching, committing, pushing, and PR creation. They will perform those actions after you exit. Do not pre-empt them even if the working tree looks ready.
 

From e7e83b0d0b299f99c4f8881db5e8474dd0afd25e Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 12:25:30 +0500
Subject: [PATCH 37/82] fix(ci): reject empty changelog file before commit
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Review feedback on PR #2460: the existing `[ -f ]` check catches
missing files but a zero-byte `docs/changelogs/v${VERSION}.md` would
still be staged and committed — `git add` + `git commit -s` on a new
empty file succeeds and produces a real commit, leaving the
downstream PR with no actual changelog content.

Add a `[ -s ]` guard after the existence check: if the Generate
changelog using AI step produces an empty file, emit a matching
`::error::` annotation and exit 1 before snapshotting.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/tags.yaml | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/.github/workflows/tags.yaml b/.github/workflows/tags.yaml
index 0ec0217d..ae77d015 100644
--- a/.github/workflows/tags.yaml
+++ b/.github/workflows/tags.yaml
@@ -328,6 +328,10 @@ jobs:
             echo "::error::Changelog file $CHANGELOG_FILE was not produced by the Generate changelog using AI step"
             exit 1
           fi
+          if [ ! -s "$CHANGELOG_FILE" ]; then
+            echo "::error::Changelog file $CHANGELOG_FILE is empty"
+            exit 1
+          fi
 
           # Snapshot the file across the branch switch — the checkout below
           # resets tracked files to match origin/main.

From 76c4eabdff8df513cca04c2313928c707df1c6ca Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 12:26:14 +0500
Subject: [PATCH 38/82] fix(ci): use a read-only app token for the Copilot step
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Review feedback on PR #2460: the Generate changelog using AI step ran
Copilot with --allow-all-tools and GH_TOKEN set to the write-capable
installation token issued to the job (contents: write,
pull-requests: write on all cozystack/* repos). The scope rules in
docs/agents/changelog.md and the step prompt tell the agent not to
use those permissions, but nothing at the token layer prevented it.

Mint a second, read-only installation token from the same app
(same COZYSTACK_CI_APP_ID / COZYSTACK_CI_PRIVATE_KEY, scoped to
contents/pull-requests/metadata read) and pass that one to the AI
step instead. The write-capable token is still used by the checkout,
commit/push, and PR-creation steps that actually need it.

This is defense in depth: even if a future prompt change or agent
misbehavior ignored the scope rules, the token itself has no write
capability on any repository in the cozystack org. No new secret,
no new GitHub App install, no admin-side change — the RO token is
minted in the same workflow from the same app credentials.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/tags.yaml | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/tags.yaml b/.github/workflows/tags.yaml
index ae77d015..90f04829 100644
--- a/.github/workflows/tags.yaml
+++ b/.github/workflows/tags.yaml
@@ -255,6 +255,21 @@ jobs:
           private-key: ${{ secrets.COZYSTACK_CI_PRIVATE_KEY }}
           owner: cozystack
 
+      # Read-only token for the AI step. Minting a separate scoped token
+      # means the Generate changelog using AI step cannot push branches,
+      # open PRs, or mutate any repository even with --allow-all-tools,
+      # regardless of whether the agent follows the prompt's instructions.
+      - name: Generate read-only GitHub App token
+        id: app-token-read
+        uses: actions/create-github-app-token@v1
+        with:
+          app-id: ${{ secrets.COZYSTACK_CI_APP_ID }}
+          private-key: ${{ secrets.COZYSTACK_CI_PRIVATE_KEY }}
+          owner: cozystack
+          permission-contents: read
+          permission-pull-requests: read
+          permission-metadata: read
+
       - name: Parse tag
         id: tag
         uses: actions/github-script@v7
@@ -306,7 +321,7 @@ jobs:
         timeout-minutes: 30
         env:
           COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_GITHUB_TOKEN }}
-          GH_TOKEN: ${{ steps.app-token.outputs.token }}
+          GH_TOKEN: ${{ steps.app-token-read.outputs.token }}
           VERSION: ${{ steps.tag.outputs.version }}
         run: |
           copilot \

From c1508940bda27b2979639981fd25461c9d4824b3 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 13:34:21 +0500
Subject: [PATCH 39/82] fix(etcd): remove destructive post-upgrade
 cert-regeneration hook

The etcd chart shipped a `post-upgrade` Helm hook that `kubectl delete`d the
etcd TLS chain (`etcd-{ca,peer-ca,client,peer,server}-tls`) and then deleted
etcd pods on every chart upgrade, gated by a semver compare of an
`etcd-deployed-version` ConfigMap against `2.6.1`.

The hook was added as a one-shot migration for the chart `2.6.0 -> 2.6.1`
transition. Since commit f871fbdb ("Remove versions_map logic") all chart
versions are stamped as `0.0.0+<git-hash>`, which per semver is always
`< 2.6.1`. The gate therefore always resolves to "update certs", firing the
destructive hook on every etcd upgrade. On clusters running Kamaji-managed
tenant control planes this wipes the etcd CA, cert-manager re-issues it, and
tenant kube-apiservers hit `x509: certificate signed by unknown authority`
against `etcd.<ns>.svc:2379` until each tenant DataStore is manually
re-reconciled.

Commit 47d81f70 ("Disabled private key rotation in CA certs") already fixed
the underlying `rotationPolicy: Always` issue the migration was papering
over, so the hook has no remaining use. Remove the hook Job, its RBAC, the
version ConfigMap it read, and add a helm-unittest suite under
`packages/extra/etcd/tests/` that guards against re-introducing the hook
or the version ConfigMap.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 packages/extra/etcd/Makefile                  |  3 ++
 packages/extra/etcd/templates/hook/job.yaml   | 39 -------------------
 packages/extra/etcd/templates/hook/role.yaml  | 26 -------------
 .../etcd/templates/hook/rolebinding.yaml      | 15 -------
 .../etcd/templates/hook/serviceaccount.yaml   |  7 ----
 packages/extra/etcd/templates/version.yaml    |  6 ---
 .../etcd/tests/no-post-upgrade-hook_test.yaml | 34 ++++++++++++++++
 7 files changed, 37 insertions(+), 93 deletions(-)
 delete mode 100644 packages/extra/etcd/templates/hook/job.yaml
 delete mode 100644 packages/extra/etcd/templates/hook/role.yaml
 delete mode 100644 packages/extra/etcd/templates/hook/rolebinding.yaml
 delete mode 100644 packages/extra/etcd/templates/hook/serviceaccount.yaml
 delete mode 100644 packages/extra/etcd/templates/version.yaml
 create mode 100644 packages/extra/etcd/tests/no-post-upgrade-hook_test.yaml

diff --git a/packages/extra/etcd/Makefile b/packages/extra/etcd/Makefile
index f37d6e1e..2b3ed61e 100644
--- a/packages/extra/etcd/Makefile
+++ b/packages/extra/etcd/Makefile
@@ -5,3 +5,6 @@ include ../../../hack/package.mk
 generate:
 	cozyvalues-gen -v values.yaml -s values.schema.json -r README.md
 	../../../hack/update-crd.sh
+
+test:
+	helm unittest .
diff --git a/packages/extra/etcd/templates/hook/job.yaml b/packages/extra/etcd/templates/hook/job.yaml
deleted file mode 100644
index 3bf2f84b..00000000
--- a/packages/extra/etcd/templates/hook/job.yaml
+++ /dev/null
@@ -1,39 +0,0 @@
-{{- $shouldUpdateCerts := true }}
-{{- $configMap := lookup "v1" "ConfigMap" .Release.Namespace "etcd-deployed-version" }}
-{{- if $configMap }}
-  {{- $deployedVersion := index $configMap "data" "version" }}
-  {{- if $deployedVersion | semverCompare ">= 2.6.1" }}
-    {{- $shouldUpdateCerts = false }}
-  {{- end }}
-{{- end }}
-
-{{- if $shouldUpdateCerts }}
----
-apiVersion: batch/v1
-kind: Job
-metadata:
-  name: etcd-hook
-  annotations:
-    helm.sh/hook: post-upgrade
-    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
-spec:
-  template:
-    metadata:
-      labels:
-        policy.cozystack.io/allow-to-apiserver: "true"
-    spec:
-      serviceAccountName: etcd-hook
-      containers:
-        - name: kubectl
-          image: docker.io/alpine/k8s:1.33.4
-          command:
-          - sh
-          args:
-            - -exc
-            - |-
-              kubectl --namespace={{ .Release.Namespace }} delete secrets etcd-ca-tls etcd-peer-ca-tls 
-              sleep 10
-              kubectl --namespace={{ .Release.Namespace }} delete secrets etcd-client-tls etcd-peer-tls etcd-server-tls
-              kubectl --namespace={{ .Release.Namespace }} delete pods --selector=app.kubernetes.io/instance=etcd,app.kubernetes.io/managed-by=etcd-operator,app.kubernetes.io/name=etcd,cozystack.io/service=etcd
-      restartPolicy: Never
-{{- end }}
diff --git a/packages/extra/etcd/templates/hook/role.yaml b/packages/extra/etcd/templates/hook/role.yaml
deleted file mode 100644
index 327eeadb..00000000
--- a/packages/extra/etcd/templates/hook/role.yaml
+++ /dev/null
@@ -1,26 +0,0 @@
-apiVersion: rbac.authorization.k8s.io/v1
-kind: Role
-metadata:
-  annotations:
-    helm.sh/hook: post-upgrade
-    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
-  name: etcd-hook
-rules:
-- apiGroups:
-  - ""
-  resources:
-  - secrets
-  - pods
-  verbs:
-  - get
-  - list
-  - watch
-  - delete
-- apiGroups:
-  - ""
-  resources:
-  - configmaps
-  verbs:
-  - get
-  - list
-  - watch
diff --git a/packages/extra/etcd/templates/hook/rolebinding.yaml b/packages/extra/etcd/templates/hook/rolebinding.yaml
deleted file mode 100644
index 0ee0ffd1..00000000
--- a/packages/extra/etcd/templates/hook/rolebinding.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-apiVersion: rbac.authorization.k8s.io/v1
-kind: RoleBinding
-metadata:
-  name: etcd-hook
-  annotations:
-    helm.sh/hook: post-upgrade
-    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
-roleRef:
-  apiGroup: rbac.authorization.k8s.io
-  kind: Role
-  name: etcd-hook
-subjects:
-  - kind: ServiceAccount
-    name: etcd-hook
-    namespace: {{ .Release.Namespace | quote }}
diff --git a/packages/extra/etcd/templates/hook/serviceaccount.yaml b/packages/extra/etcd/templates/hook/serviceaccount.yaml
deleted file mode 100644
index 552fb5fc..00000000
--- a/packages/extra/etcd/templates/hook/serviceaccount.yaml
+++ /dev/null
@@ -1,7 +0,0 @@
-apiVersion: v1
-kind: ServiceAccount
-metadata:
-  name: etcd-hook
-  annotations:
-    helm.sh/hook: post-upgrade
-    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
diff --git a/packages/extra/etcd/templates/version.yaml b/packages/extra/etcd/templates/version.yaml
deleted file mode 100644
index cc9375bb..00000000
--- a/packages/extra/etcd/templates/version.yaml
+++ /dev/null
@@ -1,6 +0,0 @@
-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: etcd-deployed-version
-data:
-  version: {{ .Chart.Version }}
diff --git a/packages/extra/etcd/tests/no-post-upgrade-hook_test.yaml b/packages/extra/etcd/tests/no-post-upgrade-hook_test.yaml
new file mode 100644
index 00000000..0e4f5aa3
--- /dev/null
+++ b/packages/extra/etcd/tests/no-post-upgrade-hook_test.yaml
@@ -0,0 +1,34 @@
+suite: etcd chart does not ship a destructive post-upgrade cert-regeneration hook
+
+release:
+  name: etcd
+  namespace: tenant-root
+
+templates:
+  - templates/check-release-name.yaml
+  - templates/dashboard-resourcemap.yaml
+  - templates/datastore.yaml
+  - templates/etcd-defrag.yaml
+  - templates/hook/job.yaml
+  - templates/podscrape.yaml
+  - templates/prometheus-rules.yaml
+  - templates/version.yaml
+
+tests:
+  - it: renders no Job named etcd-hook
+    documentSelector:
+      path: metadata.name
+      value: etcd-hook
+      skipEmptyTemplates: true
+    asserts:
+      - hasDocuments:
+          count: 0
+
+  - it: renders no ConfigMap named etcd-deployed-version
+    documentSelector:
+      path: metadata.name
+      value: etcd-deployed-version
+      skipEmptyTemplates: true
+    asserts:
+      - hasDocuments:
+          count: 0

From 9222b6feda4d97eda7481569d5d9a466b69fec94 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 14:32:20 +0500
Subject: [PATCH 40/82] ci(api): pre-fetch k8s.io/code-generator in codegen
 drift job

hack/update-codegen.sh sources kube_codegen.sh from the Go module
cache at ~/go/pkg/mod/k8s.io/code-generator@vX.Y.Z/, but the module
is not declared in go.mod so a fresh runner has nothing to source
from. Add a workflow step that parses the pinned version out of the
script and pulls the module into the cache before running
make generate.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/codegen-drift.yml | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/.github/workflows/codegen-drift.yml b/.github/workflows/codegen-drift.yml
index bd7927f7..b08f5d7e 100644
--- a/.github/workflows/codegen-drift.yml
+++ b/.github/workflows/codegen-drift.yml
@@ -30,6 +30,17 @@ jobs:
           go-version-file: go.mod
           cache: true
 
+      - name: Pre-fetch k8s.io/code-generator module
+        # hack/update-codegen.sh sources kube_codegen.sh from the Go module cache.
+        # The module is not declared in go.mod, so fetch it explicitly at the
+        # version pinned in the script.
+        run: |
+          version=$(grep -oP 'code-generator@\Kv[0-9.]+' hack/update-codegen.sh)
+          tmpdir=$(mktemp -d)
+          cd "$tmpdir"
+          go mod init codegen-fetch
+          go get "k8s.io/code-generator@${version}"
+
       - name: Run make generate
         run: make generate
 

From 0baa93006fcec09fb6245a4ccd8ec8a047448587 Mon Sep 17 00:00:00 2001
From: Kirill Ilin <stitch14@yandex.ru>
Date: Thu, 23 Apr 2026 16:57:23 +0500
Subject: [PATCH 41/82] chore(hetzner-robotlb): update robotlb chart to
 appVersion 0.0.6

Pulls the latest robotlb chart (0.1.3) which ships robotlb 0.0.6.
The new appVersion adds RBAC permissions for discovery.k8s.io/endpointslices
needed to support EndpointSlice-based services such as KubeVirt.

Assisted-By: Claude AI
Signed-off-by: Kirill Ilin <stitch14@yandex.ru>
---
 packages/system/hetzner-robotlb/charts/robotlb/Chart.yaml      | 2 +-
 .../hetzner-robotlb/charts/robotlb/templates/deployment.yaml   | 2 +-
 .../system/hetzner-robotlb/charts/robotlb/templates/role.yaml  | 3 ++-
 packages/system/hetzner-robotlb/charts/robotlb/values.yaml     | 3 +++
 4 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/packages/system/hetzner-robotlb/charts/robotlb/Chart.yaml b/packages/system/hetzner-robotlb/charts/robotlb/Chart.yaml
index 743f255d..192470dd 100644
--- a/packages/system/hetzner-robotlb/charts/robotlb/Chart.yaml
+++ b/packages/system/hetzner-robotlb/charts/robotlb/Chart.yaml
@@ -1,5 +1,5 @@
 apiVersion: v2
-appVersion: 0.0.5
+appVersion: 0.0.6
 description: A Helm chart for robotlb (loadbalancer on hetzner cloud).
 name: robotlb
 type: application
diff --git a/packages/system/hetzner-robotlb/charts/robotlb/templates/deployment.yaml b/packages/system/hetzner-robotlb/charts/robotlb/templates/deployment.yaml
index 4fd71366..41b4661a 100644
--- a/packages/system/hetzner-robotlb/charts/robotlb/templates/deployment.yaml
+++ b/packages/system/hetzner-robotlb/charts/robotlb/templates/deployment.yaml
@@ -5,7 +5,7 @@ metadata:
   labels:
     {{- include "robotlb.labels" . | nindent 4 }}
 spec:
-  replicas: {{ .Values.replicas }}
+  replicas: 1
   selector:
     matchLabels:
       {{- include "robotlb.selectorLabels" . | nindent 6 }}
diff --git a/packages/system/hetzner-robotlb/charts/robotlb/templates/role.yaml b/packages/system/hetzner-robotlb/charts/robotlb/templates/role.yaml
index 3a7b9334..76bac249 100644
--- a/packages/system/hetzner-robotlb/charts/robotlb/templates/role.yaml
+++ b/packages/system/hetzner-robotlb/charts/robotlb/templates/role.yaml
@@ -3,7 +3,8 @@ apiVersion: rbac.authorization.k8s.io/v1
 kind: ClusterRole
 metadata:
   name: {{ include "robotlb.fullname" . }}-cr
-rules: {{- toYaml .Values.serviceAccount.permissions | nindent 2 }}
+rules:
+ {{- toYaml .Values.serviceAccount.permissions | nindent 2 }}
 ---
 apiVersion: rbac.authorization.k8s.io/v1
 kind: ClusterRoleBinding
diff --git a/packages/system/hetzner-robotlb/charts/robotlb/values.yaml b/packages/system/hetzner-robotlb/charts/robotlb/values.yaml
index 739c7b73..4365f677 100644
--- a/packages/system/hetzner-robotlb/charts/robotlb/values.yaml
+++ b/packages/system/hetzner-robotlb/charts/robotlb/values.yaml
@@ -36,6 +36,9 @@ serviceAccount:
   - apiGroups: [""]
     resources: [nodes, pods]
     verbs: [get, list, watch]
+  - apiGroups: [discovery.k8s.io]
+    resources: [endpointslices]
+    verbs: [get, list, watch]
 
 podAnnotations: {}
 podLabels: {}

From ecd2ead5defaf25001cbd56e5878b9f4d2c2d27b Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 15:55:49 +0300
Subject: [PATCH 42/82] docs(agents): document make generate requirement before
 committing

Pre-commit CI runs make generate in every package and fails with exit
123 on any uncommitted generator output. Add explicit guidance so
agents stage regenerated README.md, values.schema.json and
packages/system/<name>-rd artifacts alongside the hand edits.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 68a584ee..9626e81e 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -8,6 +8,30 @@ Project-side conventions for commits, branches, and pull requests in Cozystack.
 - [ ] Commit is signed off with `--signoff`
 - [ ] Branch is rebased on `upstream/main` (no extra commits)
 - [ ] PR body includes description and release note
+- [ ] Ran `make generate` in every package whose `values.yaml`, `values.schema.json`, `Chart.yaml`, or `README.md` was touched, and committed the regenerated files
+
+## Regenerate Artifacts Before Committing
+
+Several files in each package are produced by `make generate` from `values.yaml` + `values.schema.json` and must stay in sync with the hand-edited sources:
+
+- `packages/(apps|extra)/<name>/README.md` — regenerated by `cozyvalues-gen` (parameter table, formatting).
+- `packages/(apps|extra)/<name>/values.schema.json` — `cozyvalues-gen` rewrites ordering and derived fields.
+- `packages/system/<name>-rd/cozyrds/<name>.yaml` — produced by `hack/update-crd.sh`, which `make generate` invokes.
+
+**Before committing edits to any of those sources**, run `make generate` inside the package and stage the full diff:
+
+```bash
+make -C packages/extra/<name> generate
+git add packages/extra/<name>/ packages/system/<name>-rd/
+```
+
+The repo's pre-commit CI job runs `make generate` in every package and then `git diff --exit-code`. Any unstaged generator output fails the job with exit code 123 and blocks the PR. Also rerun `make generate` after a `git commit --amend` if the amended change touched any of the sources above.
+
+To locate packages a WIP branch is likely to need regenerated:
+
+```bash
+git diff --name-only | xargs -n1 dirname | sort -u | grep ^packages/
+```
 
 ## Commit Format
 

From 0e4b66a70f036d446c5d61dfc92174ff6eae34ac Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 13:36:04 +0300
Subject: [PATCH 43/82] chore(cilium): bump to v1.19.3

Vendored chart refreshed via make update in packages/system/cilium.

Motivation: v1.19.2 fixes a cert-manager HTTP-01 bug on hostnames with
both HTTP and HTTPS listeners (cilium#44492, backport PR #44517). This
is a prerequisite for upcoming Gateway API work.

v1.19.3 is the latest stable release in the v1.19.x line (15 Apr 2026).

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../system/cilium/charts/cilium/Chart.yaml    |   4 +-
 .../system/cilium/charts/cilium/README.md     |  50 +++-
 .../configmap/bootstrap-config.yaml           |   9 +
 .../templates/cilium-agent/daemonset.yaml     |   6 +-
 .../cilium/templates/cilium-configmap.yaml    |  20 +-
 .../cilium-operator/clusterrole.yaml          |   9 +
 .../cilium/templates/ztunnel/daemonset.yaml   | 165 +++++++++++++
 .../cilium/templates/ztunnel/secret.yaml      |  23 ++
 .../templates/ztunnel/serviceaccount.yaml     |  21 ++
 .../cilium/charts/cilium/values.schema.json   | 219 +++++++++++++++++-
 .../system/cilium/charts/cilium/values.yaml   | 147 ++++++++++--
 .../cilium/charts/cilium/values.yaml.tmpl     | 106 ++++++++-
 .../system/cilium/images/cilium/Dockerfile    |   2 +-
 13 files changed, 726 insertions(+), 55 deletions(-)
 create mode 100644 packages/system/cilium/charts/cilium/templates/ztunnel/daemonset.yaml
 create mode 100644 packages/system/cilium/charts/cilium/templates/ztunnel/secret.yaml
 create mode 100644 packages/system/cilium/charts/cilium/templates/ztunnel/serviceaccount.yaml

diff --git a/packages/system/cilium/charts/cilium/Chart.yaml b/packages/system/cilium/charts/cilium/Chart.yaml
index 0bb34451..5fe2baed 100644
--- a/packages/system/cilium/charts/cilium/Chart.yaml
+++ b/packages/system/cilium/charts/cilium/Chart.yaml
@@ -76,7 +76,7 @@ annotations:
     Cilium Gateway Class Config\n  description: |\n    CiliumGatewayClassConfig defines
     a configuration for Gateway API GatewayClass.\n"
 apiVersion: v2
-appVersion: 1.19.1
+appVersion: 1.19.3
 description: eBPF-based Networking, Security, and Observability
 home: https://cilium.io/
 icon: https://cdn.jsdelivr.net/gh/cilium/cilium@main/Documentation/images/logo-solo.svg
@@ -92,4 +92,4 @@ kubeVersion: '>= 1.21.0-0'
 name: cilium
 sources:
 - https://github.com/cilium/cilium
-version: 1.19.1
+version: 1.19.3
diff --git a/packages/system/cilium/charts/cilium/README.md b/packages/system/cilium/charts/cilium/README.md
index fe8aea3f..d7697c56 100644
--- a/packages/system/cilium/charts/cilium/README.md
+++ b/packages/system/cilium/charts/cilium/README.md
@@ -1,6 +1,6 @@
 # cilium
 
-![Version: 1.19.1](https://img.shields.io/badge/Version-1.19.1-informational?style=flat-square) ![AppVersion: 1.19.1](https://img.shields.io/badge/AppVersion-1.19.1-informational?style=flat-square)
+![Version: 1.19.3](https://img.shields.io/badge/Version-1.19.3-informational?style=flat-square) ![AppVersion: 1.19.3](https://img.shields.io/badge/AppVersion-1.19.3-informational?style=flat-square)
 
 Cilium is open source software for providing and transparently securing
 network connectivity and loadbalancing between application workloads such as
@@ -89,7 +89,7 @@ contributors across the globe, there is almost always someone available to help.
 | authentication.mutual.spire.install.agent.tolerations | list | `[{"effect":"NoSchedule","key":"node.kubernetes.io/not-ready"},{"effect":"NoSchedule","key":"node-role.kubernetes.io/master"},{"effect":"NoSchedule","key":"node-role.kubernetes.io/control-plane"},{"effect":"NoSchedule","key":"node.cloudprovider.kubernetes.io/uninitialized","value":"true"},{"key":"CriticalAddonsOnly","operator":"Exists"}]` | SPIRE agent tolerations configuration By default it follows the same tolerations as the agent itself to allow the Cilium agent on this node to connect to SPIRE. ref: https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/ |
 | authentication.mutual.spire.install.enabled | bool | `true` | Enable SPIRE installation. This will only take effect only if authentication.mutual.spire.enabled is true |
 | authentication.mutual.spire.install.existingNamespace | bool | `false` | SPIRE namespace already exists. Set to true if Helm should not create, manage, and import the SPIRE namespace. |
-| authentication.mutual.spire.install.initImage | object | `{"digest":"sha256:b3255e7dfbcd10cb367af0d409747d511aeb66dfac98cf30e97e87e4207dd76f","override":null,"pullPolicy":"IfNotPresent","repository":"docker.io/library/busybox","tag":"1.37.0","useDigest":true}` | init container image of SPIRE agent and server |
+| authentication.mutual.spire.install.initImage | object | `{"digest":"sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e","override":null,"pullPolicy":"IfNotPresent","repository":"docker.io/library/busybox","tag":"1.37.0","useDigest":true}` | init container image of SPIRE agent and server |
 | authentication.mutual.spire.install.namespace | string | `"cilium-spire"` | SPIRE namespace to install into |
 | authentication.mutual.spire.install.server.affinity | object | `{}` | SPIRE server affinity configuration |
 | authentication.mutual.spire.install.server.annotations | object | `{}` | SPIRE server annotations |
@@ -175,7 +175,7 @@ contributors across the globe, there is almost always someone available to help.
 | bpf.tproxy | bool | `false` | Configure the eBPF-based TPROXY (beta) to reduce reliance on iptables rules for implementing Layer 7 policy. Note this is incompatible with netkit (`bpf.datapathMode=netkit`, `bpf.datapathMode=netkit-l2`). |
 | bpf.vlanBypass | list | `[]` | Configure explicitly allowed VLAN id's for bpf logic bypass. [0] will allow all VLAN id's without any filtering. |
 | bpfClockProbe | bool | `false` | Enable BPF clock source probing for more efficient tick retrieval. |
-| certgen | object | `{"affinity":{},"annotations":{"cronJob":{},"job":{}},"cronJob":{"failedJobsHistoryLimit":1,"successfulJobsHistoryLimit":3},"extraVolumeMounts":[],"extraVolumes":[],"generateCA":true,"image":{"digest":"sha256:19921f48ee7e2295ea4dca955878a6cd8d70e6d4219d08f688e866ece9d95d4d","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/certgen","tag":"v0.3.2","useDigest":true},"nodeSelector":{},"podLabels":{},"priorityClassName":"","resources":{},"tolerations":[],"ttlSecondsAfterFinished":null}` | Configure certificate generation for Hubble integration. If hubble.tls.auto.method=cronJob, these values are used for the Kubernetes CronJob which will be scheduled regularly to (re)generate any certificates not provided manually. |
+| certgen | object | `{"affinity":{},"annotations":{"cronJob":{},"job":{}},"cronJob":{"failedJobsHistoryLimit":1,"successfulJobsHistoryLimit":3},"extraVolumeMounts":[],"extraVolumes":[],"generateCA":true,"image":{"digest":"sha256:f0c656830e856d26b24b0e144df1f8b327d3b46748d76a630514111fc365b697","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/certgen","tag":"v0.4.1","useDigest":true},"nodeSelector":{},"podLabels":{},"priorityClassName":"","resources":{},"tolerations":[],"ttlSecondsAfterFinished":null}` | Configure certificate generation for Hubble integration. If hubble.tls.auto.method=cronJob, these values are used for the Kubernetes CronJob which will be scheduled regularly to (re)generate any certificates not provided manually. |
 | certgen.affinity | object | `{}` | Affinity for certgen |
 | certgen.annotations | object | `{"cronJob":{},"job":{}}` | Annotations to be added to the hubble-certgen initial Job and CronJob |
 | certgen.cronJob.failedJobsHistoryLimit | int | `1` | The number of failed finished jobs to keep |
@@ -214,7 +214,7 @@ contributors across the globe, there is almost always someone available to help.
 | clustermesh.apiserver.extraVolumeMounts | list | `[]` | Additional clustermesh-apiserver volumeMounts. |
 | clustermesh.apiserver.extraVolumes | list | `[]` | Additional clustermesh-apiserver volumes. |
 | clustermesh.apiserver.healthPort | int | `9880` | TCP port for the clustermesh-apiserver health API. |
-| clustermesh.apiserver.image | object | `{"digest":"sha256:56d6c3dc13b50126b80ecb571707a0ea97f6db694182b9d61efd386d04e5bb28","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/clustermesh-apiserver","tag":"v1.19.1","useDigest":true}` | Clustermesh API server image. |
+| clustermesh.apiserver.image | object | `{"digest":"sha256:a8136a7615d6c6041d3aa6f2674d17beaec238170d669507ccc05328a778e2b7","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/clustermesh-apiserver","tag":"v1.19.3","useDigest":true}` | Clustermesh API server image. |
 | clustermesh.apiserver.kvstoremesh.enabled | bool | `true` | Enable KVStoreMesh. KVStoreMesh caches the information retrieved from the remote clusters in the local etcd instance (deprecated - KVStoreMesh will always be enabled once the option is removed). |
 | clustermesh.apiserver.kvstoremesh.extraArgs | list | `[]` | Additional KVStoreMesh arguments. |
 | clustermesh.apiserver.kvstoremesh.extraEnv | list | `[]` | Additional KVStoreMesh environment variables. |
@@ -340,6 +340,10 @@ contributors across the globe, there is almost always someone available to help.
 | cni.resources | object | `{"limits":{"cpu":1,"memory":"1Gi"},"requests":{"cpu":"100m","memory":"10Mi"}}` | Specifies the resources for the cni initContainer |
 | cni.uninstall | bool | `false` | Remove the CNI configuration and binary files on agent shutdown. Enable this if you're removing Cilium from the cluster. Disable this to prevent the CNI configuration file from being removed during agent upgrade, which can cause nodes to go unmanageable. |
 | commonLabels | object | `{}` | commonLabels allows users to add common labels for all Cilium resources. |
+| configDriftDetection | object | `{"driftChecker":true,"enabled":true,"ignoredKeys":[]}` | Configuration for the ConfigMap drift detection feature. When enabled, the agent continuously watches the cilium-config ConfigMap and exposes a cilium_drift_checker_config_delta Prometheus metric reporting the number of keys that differ between the ConfigMap and the agent's active settings. A non-zero value indicates that the agent has not yet applied all current ConfigMap changes and needs to be restarted. |
+| configDriftDetection.driftChecker | bool | `true` | Enable the drift checker which compares the DynamicConfig table against the agent's active settings and publishes the cilium_drift_checker_config_delta metric. |
+| configDriftDetection.enabled | bool | `true` | Enable watching of the cilium-config ConfigMap and reflecting its contents into the agent's internal DynamicConfig table. |
+| configDriftDetection.ignoredKeys | list | `[]` | List of config-map keys to ignore when computing the drift delta. |
 | connectivityProbeFrequencyRatio | float64 | `0.5` | Ratio of the connectivity probe frequency vs resource usage, a float in [0, 1]. 0 will give more frequent probing, 1 will give less frequent probing. Probing frequency is dynamically adjusted based on the cluster size. |
 | conntrackGCInterval | string | `"0s"` | Configure how frequently garbage collection should occur for the datapath connection tracking table. |
 | conntrackGCMaxInterval | string | `""` | Configure the maximum frequency for the garbage collection of the connection tracking table. Only affects the automatic computation for the frequency and has no effect when 'conntrackGCInterval' is set. This can be set to more frequently clean up unused identities created from ToFQDN policies. |
@@ -380,7 +384,6 @@ contributors across the globe, there is almost always someone available to help.
 | enableMasqueradeRouteSource | bool | `false` | Enables masquerading to the source of the route for traffic leaving the node from endpoints. |
 | enableNoServiceEndpointsRoutable | bool | `true` | Enable routing to a service that has zero endpoints |
 | enableNonDefaultDenyPolicies | bool | `true` | Enable Non-Default-Deny policies |
-| enableTunnelBIGTCP | bool | `false` | Enable BIG TCP in tunneling mode and increase maximum GRO/GSO limits for VXLAN/GENEVE tunnels |
 | enableXTSocketFallback | bool | `true` | Enables the fallback compatibility solution for when the xt_socket kernel module is missing and it is needed for the datapath L7 redirection to work properly. See documentation for details on when this can be disabled: https://docs.cilium.io/en/stable/operations/system_requirements/#linux-kernel. |
 | encryption.enabled | bool | `false` | Enable transparent network encryption. |
 | encryption.ipsec.encryptedOverlay | bool | `false` | Enable IPsec encrypted overlay |
@@ -401,8 +404,29 @@ contributors across the globe, there is almost always someone available to help.
 | encryption.strictMode.ingress.enabled | bool | `false` | Enable strict ingress encryption. When enabled, all unencrypted overlay ingress traffic will be dropped. This option is only applicable when WireGuard and tunneling are enabled. |
 | encryption.type | string | `"ipsec"` | Encryption method. Can be one of ipsec, wireguard or ztunnel. |
 | encryption.wireguard.persistentKeepalive | string | `"0s"` | Controls WireGuard PersistentKeepalive option. Set 0s to disable. |
+| encryption.ztunnel | object | `{"affinity":{},"annotations":{},"caAddress":"https://localhost:15012","extraEnv":[],"extraVolumeMounts":[],"extraVolumes":[],"healthPort":15021,"image":{"digest":null,"override":null,"pullPolicy":"IfNotPresent","repository":"docker.io/istio/ztunnel","tag":"1.28.0-distroless","useDigest":false},"nodeSelector":{"kubernetes.io/os":"linux"},"podAnnotations":{},"podLabels":{},"priorityClassName":null,"readinessProbe":{"failureThreshold":3,"initialDelaySeconds":0,"periodSeconds":10},"resources":{"requests":{"cpu":"200m","memory":"512Mi"}},"secrets":{"bootstrapRootCert":null},"terminationGracePeriodSeconds":30,"tolerations":[{"effect":"NoSchedule","operator":"Exists"},{"key":"CriticalAddonsOnly","operator":"Exists"},{"effect":"NoExecute","operator":"Exists"}],"updateStrategy":{"rollingUpdate":{"maxSurge":1,"maxUnavailable":0},"type":"RollingUpdate"}}` | ztunnel encryption configuration. ztunnel is Istio's purpose-built, per-node proxy for handling L4 traffic in ambient mesh mode. These settings only apply when encryption.type is set to "ztunnel". |
+| encryption.ztunnel.affinity | object | `{}` | Affinity for ztunnel pods. |
+| encryption.ztunnel.annotations | object | `{}` | Annotations to be added to all ztunnel resources. |
+| encryption.ztunnel.caAddress | string | `"https://localhost:15012"` | CA server address for certificate requests. |
+| encryption.ztunnel.extraEnv | list | `[]` | Additional ztunnel container environment variables. |
+| encryption.ztunnel.extraVolumeMounts | list | `[]` | Additional ztunnel volumeMounts. |
+| encryption.ztunnel.extraVolumes | list | `[]` | Additional ztunnel volumes. |
+| encryption.ztunnel.healthPort | int | `15021` | TCP port for the health API. |
+| encryption.ztunnel.image | object | `{"digest":null,"override":null,"pullPolicy":"IfNotPresent","repository":"docker.io/istio/ztunnel","tag":"1.28.0-distroless","useDigest":false}` | ztunnel container image. |
+| encryption.ztunnel.nodeSelector | object | `{"kubernetes.io/os":"linux"}` | Node selector for ztunnel pods. |
+| encryption.ztunnel.podAnnotations | object | `{}` | Annotations to be added to ztunnel pods. |
+| encryption.ztunnel.podLabels | object | `{}` | Labels to be added to ztunnel pods. |
+| encryption.ztunnel.priorityClassName | string | `nil` | The priority class to use for ztunnel pods. |
+| encryption.ztunnel.readinessProbe | object | `{"failureThreshold":3,"initialDelaySeconds":0,"periodSeconds":10}` | Readiness probe configuration. |
+| encryption.ztunnel.resources | object | `{"requests":{"cpu":"200m","memory":"512Mi"}}` | ztunnel resource limits & requests. |
+| encryption.ztunnel.secrets | object | `{"bootstrapRootCert":null}` | ztunnel secrets configuration. |
+| encryption.ztunnel.secrets.bootstrapRootCert | string | `nil` | Base64-encoded bootstrap root certificate content. If not provided, the secret must be created manually before deploying. @schema type: [null, string] @schema |
+| encryption.ztunnel.terminationGracePeriodSeconds | int | `30` | Configure termination grace period for ztunnel DaemonSet. |
+| encryption.ztunnel.tolerations | list | `[{"effect":"NoSchedule","operator":"Exists"},{"key":"CriticalAddonsOnly","operator":"Exists"},{"effect":"NoExecute","operator":"Exists"}]` | Node tolerations for ztunnel scheduling. |
+| encryption.ztunnel.updateStrategy | object | `{"rollingUpdate":{"maxSurge":1,"maxUnavailable":0},"type":"RollingUpdate"}` | ztunnel update strategy. |
 | endpointHealthChecking.enabled | bool | `true` | Enable connectivity health checking between virtual endpoints. |
 | endpointLockdownOnMapOverflow | bool | `false` | Enable endpoint lockdown on policy map overflow. |
+| endpointPolicyUpdateTimeoutDuration | string | `nil` | Max duration to wait for envoy to respond to configuration changes. Default "10s". |
 | endpointRoutes.enabled | bool | `false` | Enable use of per endpoint routes instead of routing via the cilium_host interface. |
 | eni.awsEnablePrefixDelegation | bool | `false` | Enable ENI prefix delegation |
 | eni.awsReleaseExcessIPs | bool | `false` | Release IPs not used from the ENI |
@@ -446,7 +470,7 @@ contributors across the globe, there is almost always someone available to help.
 | envoy.httpRetryCount | int | `3` | Maximum number of retries for each HTTP request |
 | envoy.httpUpstreamLingerTimeout | string | `nil` | Time in seconds to block Envoy worker thread while an upstream HTTP connection is closing. If set to 0, the connection is closed immediately (with TCP RST). If set to -1, the connection is closed asynchronously in the background. |
 | envoy.idleTimeoutDurationSeconds | int | `60` | Set Envoy upstream HTTP idle connection timeout seconds. Does not apply to connections with pending requests. Default 60s |
-| envoy.image | object | `{"digest":"sha256:8188114a2768b5f49d6ce58e168b20d765e0fbc64eee0d83241aa2b150ccd788","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium-envoy","tag":"v1.35.9-1770979049-232ed4a26881e4ab4f766f251f258ed424fff663","useDigest":true}` | Envoy container image. |
+| envoy.image | object | `{"digest":"sha256:ba0ab8adac082d50d525fd2c5ba096c8facea3a471561b7c61c7a5b9c2e0de0d","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium-envoy","tag":"v1.36.6-1776000132-2437d2edeaf4d9b56ef279bd0d71127440c067aa","useDigest":true}` | Envoy container image. |
 | envoy.initContainers | list | `[]` | Init containers added to the cilium Envoy DaemonSet. |
 | envoy.initialFetchTimeoutSeconds | int | `30` | Time in seconds after which the initial fetch on an xDS stream is considered timed out |
 | envoy.livenessProbe.enabled | bool | `true` | Enable liveness probe for cilium-envoy |
@@ -591,7 +615,7 @@ contributors across the globe, there is almost always someone available to help.
 | hubble.relay.extraVolumes | list | `[]` | Additional hubble-relay volumes. |
 | hubble.relay.gops.enabled | bool | `true` | Enable gops for hubble-relay |
 | hubble.relay.gops.port | int | `9893` | Configure gops listen port for hubble-relay |
-| hubble.relay.image | object | `{"digest":"sha256:d8c4e13bc36a56179292bb52bc6255379cb94cb873700d316ea3139b1bdb8165","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/hubble-relay","tag":"v1.19.1","useDigest":true}` | Hubble-relay container image. |
+| hubble.relay.image | object | `{"digest":"sha256:5ee21d57b6ef2aa6db67e603a735fdceb162454b352b7335b651456e308f681b","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/hubble-relay","tag":"v1.19.3","useDigest":true}` | Hubble-relay container image. |
 | hubble.relay.listenHost | string | `""` | Host to listen to. Specify an empty string to bind to all the interfaces. |
 | hubble.relay.listenPort | string | `"4245"` | Port to listen to. |
 | hubble.relay.logOptions | object | `{"format":null,"level":null}` | Logging configuration for hubble-relay. |
@@ -709,7 +733,7 @@ contributors across the globe, there is almost always someone available to help.
 | identityAllocationMode | string | `"crd"` | Method to use for identity allocation (`crd`, `kvstore` or `doublewrite-readkvstore` / `doublewrite-readcrd` for migrating between identity backends). |
 | identityChangeGracePeriod | string | `"5s"` | Time to wait before using new identity on endpoint identity change. |
 | identityManagementMode | string | `"agent"` | Control whether CiliumIdentities are created by the agent ("agent"), the operator ("operator") or both ("both"). "Both" should be used only to migrate between "agent" and "operator". Operator-managed identities is a beta feature. |
-| image | object | `{"digest":"sha256:41f1f74a0000de8656f1de4088ea00c8f2d49d6edea579034c73c5fd5fe01792","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium","tag":"v1.19.1","useDigest":true}` | Agent container image. |
+| image | object | `{"digest":"sha256:2e61680593cddca8b6c055f6d4c849d87a26a1c91c7e3b8b56c7fb76ab7b7b10","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium","tag":"v1.19.3","useDigest":true}` | Agent container image. |
 | imagePullSecrets | list | `[]` | Configure image pull secrets for pulling container images |
 | ingressController.default | bool | `false` | Set cilium ingress controller to be the default ingress controller This will let cilium ingress controller route entries without ingress class set |
 | ingressController.defaultSecretName | string | `nil` | Default secret name for ingresses without .spec.tls[].secretName set. |
@@ -797,12 +821,13 @@ contributors across the globe, there is almost always someone available to help.
 | livenessProbe.failureThreshold | int | `10` | failure threshold of liveness probe |
 | livenessProbe.periodSeconds | int | `30` | interval between checks of the liveness probe |
 | livenessProbe.requireK8sConnectivity | bool | `false` | whether to require k8s connectivity as part of the check. |
-| loadBalancer | object | `{"acceleration":"disabled","l7":{"algorithm":"round_robin","backend":"disabled","ports":[]}}` | Configure service load balancing |
+| loadBalancer | object | `{"acceleration":"disabled","l7":{"algorithm":"round_robin","backend":"disabled","ports":[]},"serviceTopology":false}` | Configure service load balancing |
 | loadBalancer.acceleration | string | `"disabled"` | acceleration is the option to accelerate service handling via XDP Applicable values can be: disabled (do not use XDP), native (XDP BPF program is run directly out of the networking driver's early receive path), or best-effort (use native mode XDP acceleration on devices that support it). |
 | loadBalancer.l7 | object | `{"algorithm":"round_robin","backend":"disabled","ports":[]}` | L7 LoadBalancer |
 | loadBalancer.l7.algorithm | string | `"round_robin"` | Default LB algorithm The default LB algorithm to be used for services, which can be overridden by the service annotation (e.g. service.cilium.io/lb-l7-algorithm) Applicable values: round_robin, least_request, random |
 | loadBalancer.l7.backend | string | `"disabled"` | Enable L7 service load balancing via envoy proxy. The request to a k8s service, which has specific annotation e.g. service.cilium.io/lb-l7, will be forwarded to the local backend proxy to be load balanced to the service endpoints. Please refer to docs for supported annotations for more configuration.  Applicable values:   - envoy: Enable L7 load balancing via envoy proxy. This will automatically set enable-envoy-config as well.   - disabled: Disable L7 load balancing by way of service annotation. |
 | loadBalancer.l7.ports | list | `[]` | List of ports from service to be automatically redirected to above backend. Any service exposing one of these ports will be automatically redirected. Fine-grained control can be achieved by using the service annotation. |
+| loadBalancer.serviceTopology | bool | `false` | serviceTopology enables K8s Topology Aware Hints -based service endpoints filtering |
 | localRedirectPolicies.addressMatcherCIDRs | string | `nil` | Limit the allowed addresses in Address Matcher rule of Local Redirect Policies to the given CIDRs. @schema@ type: [null, array] @schema@ |
 | localRedirectPolicies.enabled | bool | `false` | Enable local redirect policies. |
 | localRedirectPolicy | bool | `false` | Enable Local Redirect Policy (deprecated, please use 'localRedirectPolicies.enabled' instead) |
@@ -860,7 +885,7 @@ contributors across the globe, there is almost always someone available to help.
 | operator.hostNetwork | bool | `true` | HostNetwork setting |
 | operator.identityGCInterval | string | `"15m0s"` | Interval for identity garbage collection. |
 | operator.identityHeartbeatTimeout | string | `"30m0s"` | Timeout for identity heartbeats. |
-| operator.image | object | `{"alibabacloudDigest":"sha256:837b12f4239e88ea5b4b5708ab982c319a94ee05edaecaafe5fd0e5b1962f554","awsDigest":"sha256:18913d05a6c4d205f0b7126c4723bb9ccbd4dc24403da46ed0f9f4bf2a142804","azureDigest":"sha256:82bce78603056e709d4c4e9f9ebb25c222c36d8a07f8c05381c2372d9078eca8","genericDigest":"sha256:e7278d763e448bf6c184b0682cf98cdca078d58a27e1b2f3c906792670aa211a","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/operator","suffix":"","tag":"v1.19.1","useDigest":true}` | cilium-operator image. |
+| operator.image | object | `{"alibabacloudDigest":"sha256:176321a65123373ff8c7823b25183102cbad98375e8d6c80b96d68b6e8491103","awsDigest":"sha256:a53dcbfb77282bf2ddd3abbe60f6d49762e7c1389a36cb35b71d504644a56640","azureDigest":"sha256:699c1571a3df1a98882ee13610d47cffb7b34ee7e8d276096db798a5f6c7e4cb","genericDigest":"sha256:205b09b0ed6accbf9fe688d312a9f0fcfc6a316fc081c23fbffb472af5dd62cd","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/operator","suffix":"","tag":"v1.19.3","useDigest":true}` | cilium-operator image. |
 | operator.nodeGCInterval | string | `"5m0s"` | Interval for cilium node garbage collection. |
 | operator.nodeSelector | object | `{"kubernetes.io/os":"linux"}` | Node labels for cilium-operator pod assignment ref: https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#nodeselector |
 | operator.podAnnotations | object | `{}` | Annotations to be added to cilium-operator pods |
@@ -918,11 +943,11 @@ contributors across the globe, there is almost always someone available to help.
 | preflight.affinity | object | `{"podAffinity":{"requiredDuringSchedulingIgnoredDuringExecution":[{"labelSelector":{"matchLabels":{"k8s-app":"cilium"}},"topologyKey":"kubernetes.io/hostname"}]}}` | Affinity for cilium-preflight |
 | preflight.annotations | object | `{}` | Annotations to be added to all top-level preflight objects (resources under templates/cilium-preflight) |
 | preflight.enabled | bool | `false` | Enable Cilium pre-flight resources (required for upgrade) |
-| preflight.envoy.image | object | `{"digest":"sha256:8188114a2768b5f49d6ce58e168b20d765e0fbc64eee0d83241aa2b150ccd788","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium-envoy","tag":"v1.35.9-1770979049-232ed4a26881e4ab4f766f251f258ed424fff663","useDigest":true}` | Envoy pre-flight image. |
+| preflight.envoy.image | object | `{"digest":"sha256:ba0ab8adac082d50d525fd2c5ba096c8facea3a471561b7c61c7a5b9c2e0de0d","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium-envoy","tag":"v1.36.6-1776000132-2437d2edeaf4d9b56ef279bd0d71127440c067aa","useDigest":true}` | Envoy pre-flight image. |
 | preflight.extraEnv | list | `[]` | Additional preflight environment variables. |
 | preflight.extraVolumeMounts | list | `[]` | Additional preflight volumeMounts. |
 | preflight.extraVolumes | list | `[]` | Additional preflight volumes. |
-| preflight.image | object | `{"digest":"sha256:41f1f74a0000de8656f1de4088ea00c8f2d49d6edea579034c73c5fd5fe01792","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium","tag":"v1.19.1","useDigest":true}` | Cilium pre-flight image. |
+| preflight.image | object | `{"digest":"sha256:2e61680593cddca8b6c055f6d4c849d87a26a1c91c7e3b8b56c7fb76ab7b7b10","override":null,"pullPolicy":"IfNotPresent","repository":"quay.io/cilium/cilium","tag":"v1.19.3","useDigest":true}` | Cilium pre-flight image. |
 | preflight.nodeSelector | object | `{"kubernetes.io/os":"linux"}` | Node labels for preflight pod assignment ref: https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#nodeselector |
 | preflight.podAnnotations | object | `{}` | Annotations to be added to preflight pods |
 | preflight.podDisruptionBudget.enabled | bool | `false` | enable PodDisruptionBudget ref: https://kubernetes.io/docs/concepts/workloads/pods/disruptions/ |
@@ -979,6 +1004,7 @@ contributors across the globe, there is almost always someone available to help.
 | serviceAccounts.corednsMCSAPI | object | `{"annotations":{},"automount":true,"create":true,"name":"cilium-coredns-mcsapi-autoconfig"}` | CorednsMCSAPI is used if clustermesh.mcsapi.corednsAutoConfigure.enabled=true |
 | serviceAccounts.hubblecertgen | object | `{"annotations":{},"automount":true,"create":true,"name":"hubble-generate-certs"}` | Hubblecertgen is used if hubble.tls.auto.method=cronJob |
 | serviceAccounts.nodeinit.enabled | bool | `false` | Enabled is temporary until https://github.com/cilium/cilium-cli/issues/1396 is implemented. Cilium CLI doesn't create the SAs for node-init, thus the workaround. Helm is not affected by this issue. Name and automount can be configured, if enabled is set to true. Otherwise, they are ignored. Enabled can be removed once the issue is fixed. Cilium-nodeinit DS must also be fixed. |
+| serviceAccounts.ztunnel | object | `{"annotations":{},"automount":false,"create":true,"name":"ztunnel-cilium"}` | Ztunnel is used if encryption.type=ztunnel |
 | serviceNoBackendResponse | string | `"reject"` | Configure what the response should be to traffic for a service without backends. Possible values:  - reject (default)  - drop |
 | sleepAfterInit | bool | `false` | Do not run Cilium agent when running with clean mode. Useful to completely uninstall Cilium as it will stop Cilium from starting and create artifacts in the node. |
 | socketLB | object | `{"enabled":false}` | Configure socket LB |
diff --git a/packages/system/cilium/charts/cilium/files/cilium-envoy/configmap/bootstrap-config.yaml b/packages/system/cilium/charts/cilium/files/cilium-envoy/configmap/bootstrap-config.yaml
index ea1d3bda..90ebf4ac 100644
--- a/packages/system/cilium/charts/cilium/files/cilium-envoy/configmap/bootstrap-config.yaml
+++ b/packages/system/cilium/charts/cilium/files/cilium-envoy/configmap/bootstrap-config.yaml
@@ -167,6 +167,8 @@ staticResources:
     circuitBreakers:
       thresholds:
       - maxRetries: {{ .Values.envoy.maxConcurrentRetries }}
+        maxConnections: {{ .Values.envoy.clusterMaxConnections }}
+        maxRequests: {{ .Values.envoy.clusterMaxRequests }}
     lbPolicy: "CLUSTER_PROVIDED"
     typedExtensionProtocolOptions:
       envoy.extensions.upstreams.http.v3.HttpProtocolOptions:
@@ -183,6 +185,8 @@ staticResources:
     circuitBreakers:
       thresholds:
       - maxRetries: {{ .Values.envoy.maxConcurrentRetries }}
+        maxConnections: {{ .Values.envoy.clusterMaxConnections }}
+        maxRequests: {{ .Values.envoy.clusterMaxRequests }}
     lbPolicy: "CLUSTER_PROVIDED"
     typedExtensionProtocolOptions:
       envoy.extensions.upstreams.http.v3.HttpProtocolOptions:
@@ -204,6 +208,8 @@ staticResources:
     circuitBreakers:
       thresholds:
       - maxRetries: {{ .Values.envoy.maxConcurrentRetries }}
+        maxConnections: {{ .Values.envoy.clusterMaxConnections }}
+        maxRequests: {{ .Values.envoy.clusterMaxRequests }}
     lbPolicy: "CLUSTER_PROVIDED"
     typedExtensionProtocolOptions:
       envoy.extensions.upstreams.http.v3.HttpProtocolOptions:
@@ -220,6 +226,8 @@ staticResources:
     circuitBreakers:
       thresholds:
       - maxRetries: {{ .Values.envoy.maxConcurrentRetries }}
+        maxConnections: {{ .Values.envoy.clusterMaxConnections }}
+        maxRequests: {{ .Values.envoy.clusterMaxRequests }}
     lbPolicy: "CLUSTER_PROVIDED"
     typedExtensionProtocolOptions:
       envoy.extensions.upstreams.http.v3.HttpProtocolOptions:
@@ -304,3 +312,4 @@ admin:
   address:
     pipe:
       path: "/var/run/cilium/envoy/sockets/admin.sock"
+      mode: 0660
diff --git a/packages/system/cilium/charts/cilium/templates/cilium-agent/daemonset.yaml b/packages/system/cilium/charts/cilium/templates/cilium-agent/daemonset.yaml
index fa3afc14..fff1b384 100644
--- a/packages/system/cilium/charts/cilium/templates/cilium-agent/daemonset.yaml
+++ b/packages/system/cilium/charts/cilium/templates/cilium-agent/daemonset.yaml
@@ -546,7 +546,7 @@ spec:
           {{- toYaml .Values.initResources | trim | nindent 10 }}
         {{- end }}
         command:
-        - sh
+        - bash
         - -ec
         # The statically linked Go program binary is invoked to avoid any
         # dependency on utilities like sh and mount that can be missing on certain
@@ -592,7 +592,7 @@ spec:
         - name: BIN_PATH
           value: {{ .Values.cni.binPath }}
         command:
-        - sh
+        - bash
         - -ec
         # The statically linked Go program binary is invoked to avoid any
         # dependency on utilities like sh that can be missing on certain
@@ -660,7 +660,7 @@ spec:
           {{- toYaml . | trim | nindent 10 }}
         {{- end }}
         command:
-        - sh
+        - bash
         - -c
         - |
           until test -s {{ (print "/tmp/cilium-bootstrap.d/" (.Values.nodeinit.bootstrapFile | base)) | quote }}; do
diff --git a/packages/system/cilium/charts/cilium/templates/cilium-configmap.yaml b/packages/system/cilium/charts/cilium/templates/cilium-configmap.yaml
index a3a38c09..5d76944d 100644
--- a/packages/system/cilium/charts/cilium/templates/cilium-configmap.yaml
+++ b/packages/system/cilium/charts/cilium/templates/cilium-configmap.yaml
@@ -464,6 +464,9 @@ data:
 {{- if has (kindOf .Values.bpf.policyMapPressureMetricsThreshold) (list "int64" "float64") }}
   bpf-policy-map-pressure-metrics-threshold: {{ .Values.bpf.policyMapPressureMetricsThreshold | quote }}
 {{- end }}
+{{- if .Values.endpointPolicyUpdateTimeoutDuration }}
+  endpoint-policy-update-timeout: {{ .Values.endpointPolicyUpdateTimeoutDuration | quote }}
+{{- end }}
 {{- if hasKey .Values.bpf "policyStatsMapMax" }}
   # bpf-policy-stats-map-max specifies the maximum number of entries in global
   # policy stats map
@@ -706,7 +709,6 @@ data:
   enable-ipv4-big-tcp: {{ .Values.enableIPv4BIGTCP | quote }}
   enable-ipv6-big-tcp: {{ .Values.enableIPv6BIGTCP | quote }}
   enable-ipv6-masquerade: {{ .Values.enableIPv6Masquerade | quote }}
-  enable-tunnel-big-tcp: {{ .Values.enableTunnelBIGTCP | quote }}
 
 {{- if hasKey .Values.bpf "enableTCX" }}
   enable-tcx: {{ .Values.bpf.enableTCX | quote }}
@@ -906,9 +908,9 @@ data:
 {{- end }}
 {{- if hasKey .Values.loadBalancer "serviceTopology" }}
   enable-service-topology: {{ .Values.loadBalancer.serviceTopology | quote }}
-# {{- end }}
-
 {{- end }}
+{{- end }}
+
 {{- if hasKey .Values.maglev "tableSize" }}
   bpf-lb-maglev-table-size: {{ .Values.maglev.tableSize | quote}}
 {{- end }}
@@ -1380,11 +1382,7 @@ data:
 
 {{- if .Values.operator.unmanagedPodWatcher.restart }}
   {{- $interval := .Values.operator.unmanagedPodWatcher.intervalSeconds }}
-  {{- if kindIs "float64" $interval }}
   unmanaged-pod-watcher-interval: {{ printf "%ds" (int $interval) | quote }}
-  {{- else }}
-  unmanaged-pod-watcher-interval: {{ $interval | quote }}
-  {{- end }}
 {{- else }}
   unmanaged-pod-watcher-interval: "0"
 {{- end }}
@@ -1517,6 +1515,14 @@ data:
   connectivity-probe-frequency-ratio: {{ .Values.connectivityProbeFrequencyRatio | quote }}
 {{- end }}
 
+{{- if hasKey .Values "configDriftDetection" }}
+  enable-dynamic-config: {{ .Values.configDriftDetection.enabled | quote }}
+  enable-drift-checker: {{ .Values.configDriftDetection.driftChecker | quote }}
+  {{- if .Values.configDriftDetection.ignoredKeys }}
+  ignore-flags-drift-checker: {{ join "," .Values.configDriftDetection.ignoredKeys | quote }}
+  {{- end }}
+{{- end }}
+
 # Extra config allows adding arbitrary properties to the cilium config.
 # By putting it at the end of the ConfigMap, it's also possible to override existing properties.
 {{- if .Values.extraConfig }}
diff --git a/packages/system/cilium/charts/cilium/templates/cilium-operator/clusterrole.yaml b/packages/system/cilium/charts/cilium/templates/cilium-operator/clusterrole.yaml
index 96f72c7f..c147e2b5 100644
--- a/packages/system/cilium/charts/cilium/templates/cilium-operator/clusterrole.yaml
+++ b/packages/system/cilium/charts/cilium/templates/cilium-operator/clusterrole.yaml
@@ -407,6 +407,15 @@ rules:
   verbs:
   - update
   - patch
+- apiGroups:
+  - multicluster.x-k8s.io
+  resources:
+  # The controller needs to be able to set serviceimport finalizers to be able to create a derived Service
+  # resource that is owned by the ServiceImport and sets blockOwnerDeletion=true in its ownerRef.
+  # This is required when the admission plugin OwnerReferencesPermissionEnforcement is activated.
+  - serviceimports/finalizers
+  verbs:
+  - update
 - apiGroups:
   - multicluster.x-k8s.io
   resources:
diff --git a/packages/system/cilium/charts/cilium/templates/ztunnel/daemonset.yaml b/packages/system/cilium/charts/cilium/templates/ztunnel/daemonset.yaml
new file mode 100644
index 00000000..f9363068
--- /dev/null
+++ b/packages/system/cilium/charts/cilium/templates/ztunnel/daemonset.yaml
@@ -0,0 +1,165 @@
+{{- if and .Values.encryption.enabled (eq .Values.encryption.type "ztunnel") }}
+---
+kind: DaemonSet
+apiVersion: apps/v1
+metadata:
+  name: ztunnel-cilium
+  namespace: {{ include "cilium.namespace" . }}
+  {{- with .Values.encryption.ztunnel.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+  labels:
+    app: ztunnel-cilium
+    app.kubernetes.io/part-of: cilium
+    app.kubernetes.io/name: ztunnel-cilium
+    {{- with .Values.commonLabels }}
+    {{- toYaml . | nindent 4 }}
+    {{- end }}
+spec:
+  revisionHistoryLimit: 10
+  selector:
+    matchLabels:
+      app: ztunnel-cilium
+  {{- with .Values.encryption.ztunnel.updateStrategy }}
+  updateStrategy:
+    {{- toYaml . | trim | nindent 4 }}
+  {{- end }}
+  template:
+    metadata:
+      annotations:
+        sidecar.istio.io/inject: "false"
+        {{- with .Values.encryption.ztunnel.podAnnotations }}
+        {{- toYaml . | nindent 8 }}
+        {{- end }}
+      labels:
+        app: ztunnel-cilium
+        app.kubernetes.io/part-of: cilium
+        app.kubernetes.io/name: ztunnel-cilium
+        {{- with .Values.commonLabels }}
+        {{- toYaml . | nindent 8 }}
+        {{- end }}
+        {{- with .Values.encryption.ztunnel.podLabels }}
+        {{- toYaml . | nindent 8 }}
+        {{- end }}
+    spec:
+      {{- with .Values.imagePullSecrets }}
+      imagePullSecrets:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      hostNetwork: true
+      dnsPolicy: ClusterFirst
+      containers:
+        - name: istio-proxy
+          image: {{ include "cilium.image" .Values.encryption.ztunnel.image | quote }}
+          imagePullPolicy: {{ .Values.encryption.ztunnel.image.pullPolicy }}
+          args:
+            - proxy
+            - ztunnel
+          env:
+            - name: XDS_ADDRESS
+              value: "https://localhost:15012"
+            - name: XDS_ROOT_CA
+              value: "/etc/ztunnel/bootstrap-root.crt"
+            - name: CA_ROOT_CA
+              value: "/etc/ztunnel/bootstrap-root.crt"
+            - name: CA_ADDRESS
+              value: {{ .Values.encryption.ztunnel.caAddress | quote }}
+            - name: ISTIO_META_DNS_CAPTURE
+              value: "false"
+            - name: INPOD_UDS
+              value: "/var/run/cilium/ztunnel.sock"
+            - name: NODE_NAME
+              valueFrom:
+                fieldRef:
+                  apiVersion: v1
+                  fieldPath: status.hostIP
+            - name: INSTANCE_IP
+              valueFrom:
+                fieldRef:
+                  apiVersion: v1
+                  fieldPath: status.podIP
+            - name: ISTIO_META_ENABLE_HBONE
+              value: "true"
+            {{- with .Values.encryption.ztunnel.extraEnv }}
+            {{- toYaml . | trim | nindent 12 }}
+            {{- end }}
+          readinessProbe:
+            httpGet:
+              path: /healthz/ready
+              port: {{ .Values.encryption.ztunnel.healthPort }}
+              host: "127.0.0.1"
+              scheme: HTTP
+            initialDelaySeconds: {{ .Values.encryption.ztunnel.readinessProbe.initialDelaySeconds }}
+            periodSeconds: {{ .Values.encryption.ztunnel.readinessProbe.periodSeconds }}
+            failureThreshold: {{ .Values.encryption.ztunnel.readinessProbe.failureThreshold }}
+            successThreshold: 1
+            timeoutSeconds: 1
+          {{- with .Values.encryption.ztunnel.resources }}
+          resources:
+            {{- toYaml . | trim | nindent 12 }}
+          {{- end }}
+          securityContext:
+            allowPrivilegeEscalation: true
+            capabilities:
+              add:
+                - NET_ADMIN
+                - SYS_ADMIN
+                - NET_RAW
+              drop:
+                - ALL
+            privileged: false
+            readOnlyRootFilesystem: true
+            runAsGroup: 1337
+            runAsNonRoot: false
+            runAsUser: 0
+          terminationMessagePath: /dev/termination-log
+          terminationMessagePolicy: File
+          volumeMounts:
+            - mountPath: /var/run/cilium
+              name: cilium-dir
+              readOnly: false
+            - mountPath: /etc/ztunnel
+              name: cilium-ztunnel-secrets
+              readOnly: true
+            {{- with .Values.encryption.ztunnel.extraVolumeMounts }}
+            {{- toYaml . | nindent 12 }}
+            {{- end }}
+      {{- with .Values.encryption.ztunnel.affinity }}
+      affinity:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.encryption.ztunnel.nodeSelector }}
+      nodeSelector:
+        {{- toYaml . | trim | nindent 8 }}
+      {{- end }}
+      {{- with .Values.encryption.ztunnel.tolerations }}
+      tolerations:
+        {{- toYaml . | trim | nindent 8 }}
+      {{- end }}
+      priorityClassName: {{ include "cilium.priorityClass" (list $ .Values.encryption.ztunnel.priorityClassName "system-node-critical") }}
+      restartPolicy: Always
+      terminationGracePeriodSeconds: {{ .Values.encryption.ztunnel.terminationGracePeriodSeconds }}
+      {{- if .Values.serviceAccounts.ztunnel.create }}
+      serviceAccountName: {{ .Values.serviceAccounts.ztunnel.name | quote }}
+      automountServiceAccountToken: {{ .Values.serviceAccounts.ztunnel.automount }}
+      {{- else }}
+      automountServiceAccountToken: false
+      {{- end }}
+      volumes:
+        - name: cilium-dir
+          hostPath:
+            path: /var/run/cilium
+            type: DirectoryOrCreate
+        - name: cilium-ztunnel-secrets
+          secret:
+            secretName: cilium-ztunnel-secrets
+            defaultMode: 420
+            items:
+              - key: bootstrap-root.crt
+                path: bootstrap-root.crt
+                mode: 420
+        {{- with .Values.encryption.ztunnel.extraVolumes }}
+        {{- toYaml . | nindent 8 }}
+        {{- end }}
+{{- end }}
diff --git a/packages/system/cilium/charts/cilium/templates/ztunnel/secret.yaml b/packages/system/cilium/charts/cilium/templates/ztunnel/secret.yaml
new file mode 100644
index 00000000..520857dc
--- /dev/null
+++ b/packages/system/cilium/charts/cilium/templates/ztunnel/secret.yaml
@@ -0,0 +1,23 @@
+{{- if and .Values.encryption.enabled (eq .Values.encryption.type "ztunnel") }}
+{{- if .Values.encryption.ztunnel.secrets.bootstrapRootCert }}
+---
+apiVersion: v1
+kind: Secret
+metadata:
+  name: cilium-ztunnel-secrets
+  namespace: {{ include "cilium.namespace" . }}
+  {{- with .Values.encryption.ztunnel.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+  labels:
+    app.kubernetes.io/part-of: cilium
+    app.kubernetes.io/name: ztunnel-cilium
+    {{- with .Values.commonLabels }}
+    {{- toYaml . | nindent 4 }}
+    {{- end }}
+type: Opaque
+data:
+  bootstrap-root.crt: {{ .Values.encryption.ztunnel.secrets.bootstrapRootCert | b64enc }}
+{{- end }}
+{{- end }}
diff --git a/packages/system/cilium/charts/cilium/templates/ztunnel/serviceaccount.yaml b/packages/system/cilium/charts/cilium/templates/ztunnel/serviceaccount.yaml
new file mode 100644
index 00000000..4ef9e2bc
--- /dev/null
+++ b/packages/system/cilium/charts/cilium/templates/ztunnel/serviceaccount.yaml
@@ -0,0 +1,21 @@
+{{- if and .Values.encryption.enabled (eq .Values.encryption.type "ztunnel") .Values.serviceAccounts.ztunnel.create }}
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ .Values.serviceAccounts.ztunnel.name | quote }}
+  namespace: {{ include "cilium.namespace" . }}
+  {{- with .Values.commonLabels }}
+  labels:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+  {{- if or .Values.serviceAccounts.ztunnel.annotations .Values.encryption.ztunnel.annotations }}
+  annotations:
+    {{- with .Values.encryption.ztunnel.annotations }}
+      {{- toYaml . | nindent 4 }}
+    {{- end }}
+    {{- with .Values.serviceAccounts.ztunnel.annotations }}
+      {{- toYaml . | nindent 4 }}
+    {{- end }}
+  {{- end }}
+{{- end }}
diff --git a/packages/system/cilium/charts/cilium/values.schema.json b/packages/system/cilium/charts/cilium/values.schema.json
index d18566d7..65c84af5 100644
--- a/packages/system/cilium/charts/cilium/values.schema.json
+++ b/packages/system/cilium/charts/cilium/values.schema.json
@@ -1714,6 +1714,21 @@
         "object"
       ]
     },
+    "configDriftDetection": {
+      "properties": {
+        "driftChecker": {
+          "type": "boolean"
+        },
+        "enabled": {
+          "type": "boolean"
+        },
+        "ignoredKeys": {
+          "items": {},
+          "type": "array"
+        }
+      },
+      "type": "object"
+    },
     "connectivityProbeFrequencyRatio": {
       "type": [
         "null",
@@ -1898,9 +1913,6 @@
     "enableNonDefaultDenyPolicies": {
       "type": "boolean"
     },
-    "enableTunnelBIGTCP": {
-      "type": "boolean"
-    },
     "enableXTSocketFallback": {
       "type": "boolean"
     },
@@ -1984,6 +1996,184 @@
             }
           },
           "type": "object"
+        },
+        "ztunnel": {
+          "properties": {
+            "affinity": {
+              "type": "object"
+            },
+            "annotations": {
+              "type": "object"
+            },
+            "caAddress": {
+              "type": "string"
+            },
+            "extraEnv": {
+              "items": {},
+              "type": "array"
+            },
+            "extraVolumeMounts": {
+              "items": {},
+              "type": "array"
+            },
+            "extraVolumes": {
+              "items": {},
+              "type": "array"
+            },
+            "healthPort": {
+              "type": "integer"
+            },
+            "image": {
+              "properties": {
+                "digest": {
+                  "type": [
+                    "null",
+                    "string"
+                  ]
+                },
+                "override": {
+                  "type": [
+                    "null",
+                    "string"
+                  ]
+                },
+                "pullPolicy": {
+                  "type": "string"
+                },
+                "repository": {
+                  "type": "string"
+                },
+                "tag": {
+                  "type": "string"
+                },
+                "useDigest": {
+                  "type": "boolean"
+                }
+              },
+              "type": "object"
+            },
+            "nodeSelector": {
+              "properties": {
+                "kubernetes.io/os": {
+                  "type": "string"
+                }
+              },
+              "type": "object"
+            },
+            "podAnnotations": {
+              "type": "object"
+            },
+            "podLabels": {
+              "type": "object"
+            },
+            "priorityClassName": {
+              "type": [
+                "null",
+                "string"
+              ]
+            },
+            "readinessProbe": {
+              "properties": {
+                "failureThreshold": {
+                  "type": "integer"
+                },
+                "initialDelaySeconds": {
+                  "type": "integer"
+                },
+                "periodSeconds": {
+                  "type": "integer"
+                }
+              },
+              "type": "object"
+            },
+            "resources": {
+              "properties": {
+                "requests": {
+                  "properties": {
+                    "cpu": {
+                      "type": "string"
+                    },
+                    "memory": {
+                      "type": "string"
+                    }
+                  },
+                  "type": "object"
+                }
+              },
+              "type": "object"
+            },
+            "secrets": {
+              "properties": {
+                "bootstrapRootCert": {
+                  "type": [
+                    "null",
+                    "string"
+                  ]
+                }
+              },
+              "type": "object"
+            },
+            "terminationGracePeriodSeconds": {
+              "type": "integer"
+            },
+            "tolerations": {
+              "items": {
+                "anyOf": [
+                  {
+                    "properties": {
+                      "effect": {
+                        "type": "string"
+                      },
+                      "operator": {
+                        "type": "string"
+                      }
+                    }
+                  },
+                  {
+                    "properties": {
+                      "key": {
+                        "type": "string"
+                      },
+                      "operator": {
+                        "type": "string"
+                      }
+                    }
+                  },
+                  {
+                    "properties": {
+                      "effect": {
+                        "type": "string"
+                      },
+                      "operator": {
+                        "type": "string"
+                      }
+                    }
+                  }
+                ]
+              },
+              "type": "array"
+            },
+            "updateStrategy": {
+              "properties": {
+                "rollingUpdate": {
+                  "properties": {
+                    "maxSurge": {
+                      "type": "integer"
+                    },
+                    "maxUnavailable": {
+                      "type": "integer"
+                    }
+                  },
+                  "type": "object"
+                },
+                "type": {
+                  "type": "string"
+                }
+              },
+              "type": "object"
+            }
+          },
+          "type": "object"
         }
       },
       "type": "object"
@@ -1999,6 +2189,9 @@
     "endpointLockdownOnMapOverflow": {
       "type": "boolean"
     },
+    "endpointPolicyUpdateTimeoutDuration": {
+      "type": "null"
+    },
     "endpointRoutes": {
       "properties": {
         "enabled": {
@@ -4484,6 +4677,9 @@
             }
           },
           "type": "object"
+        },
+        "serviceTopology": {
+          "type": "boolean"
         }
       },
       "type": "object"
@@ -6029,6 +6225,23 @@
             }
           },
           "type": "object"
+        },
+        "ztunnel": {
+          "properties": {
+            "annotations": {
+              "type": "object"
+            },
+            "automount": {
+              "type": "boolean"
+            },
+            "create": {
+              "type": "boolean"
+            },
+            "name": {
+              "type": "string"
+            }
+          },
+          "type": "object"
         }
       },
       "type": "object"
diff --git a/packages/system/cilium/charts/cilium/values.yaml b/packages/system/cilium/charts/cilium/values.yaml
index b9b30830..66f76982 100644
--- a/packages/system/cilium/charts/cilium/values.yaml
+++ b/packages/system/cilium/charts/cilium/values.yaml
@@ -210,6 +210,12 @@ serviceAccounts:
     name: cilium-coredns-mcsapi-autoconfig
     automount: true
     annotations: {}
+  # -- Ztunnel is used if encryption.type=ztunnel
+  ztunnel:
+    create: true
+    name: ztunnel-cilium
+    automount: false
+    annotations: {}
 # -- Configure termination grace period for cilium-agent DaemonSet.
 terminationGracePeriodSeconds: 1
 # -- Install the cilium agent resources.
@@ -218,6 +224,22 @@ agent: true
 name: cilium
 # -- Roll out cilium agent pods automatically when configmap is updated.
 rollOutCiliumPods: false
+# -- Configuration for the ConfigMap drift detection feature.
+# When enabled, the agent continuously watches the cilium-config ConfigMap
+# and exposes a cilium_drift_checker_config_delta Prometheus metric reporting
+# the number of keys that differ between the ConfigMap and the agent's active
+# settings. A non-zero value indicates that the agent has not yet applied all
+# current ConfigMap changes and needs to be restarted.
+configDriftDetection:
+  # -- Enable watching of the cilium-config ConfigMap and reflecting its
+  # contents into the agent's internal DynamicConfig table.
+  enabled: true
+  # -- Enable the drift checker which compares the DynamicConfig table against
+  # the agent's active settings and publishes the
+  # cilium_drift_checker_config_delta metric.
+  driftChecker: true
+  # -- List of config-map keys to ignore when computing the drift delta.
+  ignoredKeys: []
 # -- Agent container image.
 image:
   # @schema
@@ -225,10 +247,10 @@ image:
   # @schema
   override: ~
   repository: "quay.io/cilium/cilium"
-  tag: "v1.19.1"
+  tag: "v1.19.3"
   pullPolicy: "IfNotPresent"
   # cilium-digest
-  digest: sha256:41f1f74a0000de8656f1de4088ea00c8f2d49d6edea579034c73c5fd5fe01792
+  digest: sha256:2e61680593cddca8b6c055f6d4c849d87a26a1c91c7e3b8b56c7fb76ab7b7b10
   useDigest: true
 # -- Scheduling configurations for cilium pods
 scheduling:
@@ -1133,9 +1155,89 @@ encryption:
   wireguard:
     # -- Controls WireGuard PersistentKeepalive option. Set 0s to disable.
     persistentKeepalive: 0s
+  # -- ztunnel encryption configuration.
+  # ztunnel is Istio's purpose-built, per-node proxy for handling L4 traffic in ambient mesh mode.
+  # These settings only apply when encryption.type is set to "ztunnel".
+  ztunnel:
+    # -- ztunnel container image.
+    image:
+      # @schema
+      # type: [null, string]
+      # @schema
+      override: ~
+      repository: "docker.io/istio/ztunnel"
+      tag: "1.28.0-distroless"
+      pullPolicy: "IfNotPresent"
+      # @schema
+      # type: [null, string]
+      # @schema
+      digest: ~
+      useDigest: false
+    # -- CA server address for certificate requests.
+    caAddress: "https://localhost:15012"
+    # -- TCP port for the health API.
+    healthPort: 15021
+    # -- ztunnel resource limits & requests.
+    resources:
+      requests:
+        cpu: 200m
+        memory: 512Mi
+    # -- ztunnel update strategy.
+    updateStrategy:
+      type: RollingUpdate
+      rollingUpdate:
+        maxSurge: 1
+        maxUnavailable: 0
+    # -- Configure termination grace period for ztunnel DaemonSet.
+    terminationGracePeriodSeconds: 30
+    # -- Readiness probe configuration.
+    readinessProbe:
+      initialDelaySeconds: 0
+      periodSeconds: 10
+      failureThreshold: 3
+    # -- Node selector for ztunnel pods.
+    nodeSelector:
+      kubernetes.io/os: linux
+    # -- Node tolerations for ztunnel scheduling.
+    tolerations:
+      - effect: NoSchedule
+        operator: Exists
+      - key: CriticalAddonsOnly
+        operator: Exists
+      - effect: NoExecute
+        operator: Exists
+    # -- Affinity for ztunnel pods.
+    affinity: {}
+    # @schema
+    # type: [null, string]
+    # @schema
+    # -- The priority class to use for ztunnel pods.
+    priorityClassName: ~
+    # -- Annotations to be added to all ztunnel resources.
+    annotations: {}
+    # -- Annotations to be added to ztunnel pods.
+    podAnnotations: {}
+    # -- Labels to be added to ztunnel pods.
+    podLabels: {}
+    # -- Additional ztunnel container environment variables.
+    extraEnv: []
+    # -- Additional ztunnel volumes.
+    extraVolumes: []
+    # -- Additional ztunnel volumeMounts.
+    extraVolumeMounts: []
+    # -- ztunnel secrets configuration.
+    secrets:
+      # -- Base64-encoded bootstrap root certificate content.
+      # If not provided, the secret must be created manually before deploying.
+      # @schema
+      # type: [null, string]
+      # @schema
+      bootstrapRootCert: ~
 endpointHealthChecking:
   # -- Enable connectivity health checking between virtual endpoints.
   enabled: true
+# -- Max duration to wait for envoy to respond to configuration changes. Default "10s".
+endpointPolicyUpdateTimeoutDuration: null
 endpointRoutes:
   # @schema
   # type: [boolean, string]
@@ -1250,8 +1352,8 @@ certgen:
     # @schema
     override: ~
     repository: "quay.io/cilium/certgen"
-    tag: "v0.3.2"
-    digest: "sha256:19921f48ee7e2295ea4dca955878a6cd8d70e6d4219d08f688e866ece9d95d4d"
+    tag: "v0.4.1"
+    digest: "sha256:f0c656830e856d26b24b0e144df1f8b327d3b46748d76a630514111fc365b697"
     useDigest: true
     pullPolicy: "IfNotPresent"
   # @schema
@@ -1599,9 +1701,9 @@ hubble:
       # @schema
       override: ~
       repository: "quay.io/cilium/hubble-relay"
-      tag: "v1.19.1"
+      tag: "v1.19.3"
       # hubble-relay-digest
-      digest: sha256:d8c4e13bc36a56179292bb52bc6255379cb94cb873700d316ea3139b1bdb8165
+      digest: sha256:5ee21d57b6ef2aa6db67e603a735fdceb162454b352b7335b651456e308f681b
       useDigest: true
       pullPolicy: "IfNotPresent"
     # -- Specifies the resources for the hubble-relay pods
@@ -2313,8 +2415,6 @@ enableMasqueradeRouteSource: false
 enableIPv4BIGTCP: false
 # -- Enables IPv6 BIG TCP support which increases maximum IPv6 GSO/GRO limits for nodes and pods
 enableIPv6BIGTCP: false
-# -- Enable BIG TCP in tunneling mode and increase maximum GRO/GSO limits for VXLAN/GENEVE tunnels
-enableTunnelBIGTCP: false
 nat:
   # -- Number of the top-k SNAT map connections to track in Cilium statedb.
   mapStatsEntries: 32
@@ -2392,8 +2492,7 @@ loadBalancer:
 
   # -- serviceTopology enables K8s Topology Aware Hints -based service
   # endpoints filtering
-  # serviceTopology: false
-
+  serviceTopology: false
   # -- L7 LoadBalancer
   l7:
     # -- Enable L7 service load balancing via envoy proxy.
@@ -2626,9 +2725,9 @@ envoy:
     # @schema
     override: ~
     repository: "quay.io/cilium/cilium-envoy"
-    tag: "v1.35.9-1770979049-232ed4a26881e4ab4f766f251f258ed424fff663"
+    tag: "v1.36.6-1776000132-2437d2edeaf4d9b56ef279bd0d71127440c067aa"
     pullPolicy: "IfNotPresent"
-    digest: "sha256:8188114a2768b5f49d6ce58e168b20d765e0fbc64eee0d83241aa2b150ccd788"
+    digest: "sha256:ba0ab8adac082d50d525fd2c5ba096c8facea3a471561b7c61c7a5b9c2e0de0d"
     useDigest: true
   # -- Init containers added to the cilium Envoy DaemonSet.
   initContainers: []
@@ -3011,15 +3110,15 @@ operator:
     # @schema
     override: ~
     repository: "quay.io/cilium/operator"
-    tag: "v1.19.1"
+    tag: "v1.19.3"
     # operator-generic-digest
-    genericDigest: sha256:e7278d763e448bf6c184b0682cf98cdca078d58a27e1b2f3c906792670aa211a
+    genericDigest: sha256:205b09b0ed6accbf9fe688d312a9f0fcfc6a316fc081c23fbffb472af5dd62cd
     # operator-azure-digest
-    azureDigest: sha256:82bce78603056e709d4c4e9f9ebb25c222c36d8a07f8c05381c2372d9078eca8
+    azureDigest: sha256:699c1571a3df1a98882ee13610d47cffb7b34ee7e8d276096db798a5f6c7e4cb
     # operator-aws-digest
-    awsDigest: sha256:18913d05a6c4d205f0b7126c4723bb9ccbd4dc24403da46ed0f9f4bf2a142804
+    awsDigest: sha256:a53dcbfb77282bf2ddd3abbe60f6d49762e7c1389a36cb35b71d504644a56640
     # operator-alibabacloud-digest
-    alibabacloudDigest: sha256:837b12f4239e88ea5b4b5708ab982c319a94ee05edaecaafe5fd0e5b1962f554
+    alibabacloudDigest: sha256:176321a65123373ff8c7823b25183102cbad98375e8d6c80b96d68b6e8491103
     useDigest: true
     pullPolicy: "IfNotPresent"
     suffix: ""
@@ -3344,9 +3443,9 @@ preflight:
     # @schema
     override: ~
     repository: "quay.io/cilium/cilium"
-    tag: "v1.19.1"
+    tag: "v1.19.3"
     # cilium-digest
-    digest: sha256:41f1f74a0000de8656f1de4088ea00c8f2d49d6edea579034c73c5fd5fe01792
+    digest: sha256:2e61680593cddca8b6c055f6d4c849d87a26a1c91c7e3b8b56c7fb76ab7b7b10
     useDigest: true
     pullPolicy: "IfNotPresent"
   envoy:
@@ -3357,9 +3456,9 @@ preflight:
       # @schema
       override: ~
       repository: "quay.io/cilium/cilium-envoy"
-      tag: "v1.35.9-1770979049-232ed4a26881e4ab4f766f251f258ed424fff663"
+      tag: "v1.36.6-1776000132-2437d2edeaf4d9b56ef279bd0d71127440c067aa"
       pullPolicy: "IfNotPresent"
-      digest: "sha256:8188114a2768b5f49d6ce58e168b20d765e0fbc64eee0d83241aa2b150ccd788"
+      digest: "sha256:ba0ab8adac082d50d525fd2c5ba096c8facea3a471561b7c61c7a5b9c2e0de0d"
       useDigest: true
   # -- The priority class to use for the preflight pod.
   priorityClassName: ""
@@ -3603,9 +3702,9 @@ clustermesh:
       # @schema
       override: ~
       repository: "quay.io/cilium/clustermesh-apiserver"
-      tag: "v1.19.1"
+      tag: "v1.19.3"
       # clustermesh-apiserver-digest
-      digest: sha256:56d6c3dc13b50126b80ecb571707a0ea97f6db694182b9d61efd386d04e5bb28
+      digest: sha256:a8136a7615d6c6041d3aa6f2674d17beaec238170d669507ccc05328a778e2b7
       useDigest: true
       pullPolicy: "IfNotPresent"
     # -- TCP port for the clustermesh-apiserver health API.
@@ -4140,7 +4239,7 @@ authentication:
           override: ~
           repository: "docker.io/library/busybox"
           tag: "1.37.0"
-          digest: "sha256:b3255e7dfbcd10cb367af0d409747d511aeb66dfac98cf30e97e87e4207dd76f"
+          digest: "sha256:1487d0af5f52b4ba31c7e465126ee2123fe3f2305d638e7827681e7cf6c83d5e"
           useDigest: true
           pullPolicy: "IfNotPresent"
         # SPIRE agent configuration
diff --git a/packages/system/cilium/charts/cilium/values.yaml.tmpl b/packages/system/cilium/charts/cilium/values.yaml.tmpl
index c21f1c26..8039b40c 100644
--- a/packages/system/cilium/charts/cilium/values.yaml.tmpl
+++ b/packages/system/cilium/charts/cilium/values.yaml.tmpl
@@ -213,6 +213,12 @@ serviceAccounts:
     name: cilium-coredns-mcsapi-autoconfig
     automount: true
     annotations: {}
+  # -- Ztunnel is used if encryption.type=ztunnel
+  ztunnel:
+    create: true
+    name: ztunnel-cilium
+    automount: false
+    annotations: {}
 # -- Configure termination grace period for cilium-agent DaemonSet.
 terminationGracePeriodSeconds: 1
 # -- Install the cilium agent resources.
@@ -221,6 +227,22 @@ agent: true
 name: cilium
 # -- Roll out cilium agent pods automatically when configmap is updated.
 rollOutCiliumPods: false
+# -- Configuration for the ConfigMap drift detection feature.
+# When enabled, the agent continuously watches the cilium-config ConfigMap
+# and exposes a cilium_drift_checker_config_delta Prometheus metric reporting
+# the number of keys that differ between the ConfigMap and the agent's active
+# settings. A non-zero value indicates that the agent has not yet applied all
+# current ConfigMap changes and needs to be restarted.
+configDriftDetection:
+  # -- Enable watching of the cilium-config ConfigMap and reflecting its
+  # contents into the agent's internal DynamicConfig table.
+  enabled: true
+  # -- Enable the drift checker which compares the DynamicConfig table against
+  # the agent's active settings and publishes the
+  # cilium_drift_checker_config_delta metric.
+  driftChecker: true
+  # -- List of config-map keys to ignore when computing the drift delta.
+  ignoredKeys: []
 # -- Agent container image.
 image:
   # @schema
@@ -1148,9 +1170,89 @@ encryption:
   wireguard:
     # -- Controls WireGuard PersistentKeepalive option. Set 0s to disable.
     persistentKeepalive: 0s
+  # -- ztunnel encryption configuration.
+  # ztunnel is Istio's purpose-built, per-node proxy for handling L4 traffic in ambient mesh mode.
+  # These settings only apply when encryption.type is set to "ztunnel".
+  ztunnel:
+    # -- ztunnel container image.
+    image:
+      # @schema
+      # type: [null, string]
+      # @schema
+      override: ~
+      repository: "docker.io/istio/ztunnel"
+      tag: "1.28.0-distroless"
+      pullPolicy: "IfNotPresent"
+      # @schema
+      # type: [null, string]
+      # @schema
+      digest: ~
+      useDigest: false
+    # -- CA server address for certificate requests.
+    caAddress: "https://localhost:15012"
+    # -- TCP port for the health API.
+    healthPort: 15021
+    # -- ztunnel resource limits & requests.
+    resources:
+      requests:
+        cpu: 200m
+        memory: 512Mi
+    # -- ztunnel update strategy.
+    updateStrategy:
+      type: RollingUpdate
+      rollingUpdate:
+        maxSurge: 1
+        maxUnavailable: 0
+    # -- Configure termination grace period for ztunnel DaemonSet.
+    terminationGracePeriodSeconds: 30
+    # -- Readiness probe configuration.
+    readinessProbe:
+      initialDelaySeconds: 0
+      periodSeconds: 10
+      failureThreshold: 3
+    # -- Node selector for ztunnel pods.
+    nodeSelector:
+      kubernetes.io/os: linux
+    # -- Node tolerations for ztunnel scheduling.
+    tolerations:
+      - effect: NoSchedule
+        operator: Exists
+      - key: CriticalAddonsOnly
+        operator: Exists
+      - effect: NoExecute
+        operator: Exists
+    # -- Affinity for ztunnel pods.
+    affinity: {}
+    # @schema
+    # type: [null, string]
+    # @schema
+    # -- The priority class to use for ztunnel pods.
+    priorityClassName: ~
+    # -- Annotations to be added to all ztunnel resources.
+    annotations: {}
+    # -- Annotations to be added to ztunnel pods.
+    podAnnotations: {}
+    # -- Labels to be added to ztunnel pods.
+    podLabels: {}
+    # -- Additional ztunnel container environment variables.
+    extraEnv: []
+    # -- Additional ztunnel volumes.
+    extraVolumes: []
+    # -- Additional ztunnel volumeMounts.
+    extraVolumeMounts: []
+    # -- ztunnel secrets configuration.
+    secrets:
+      # -- Base64-encoded bootstrap root certificate content.
+      # If not provided, the secret must be created manually before deploying.
+      # @schema
+      # type: [null, string]
+      # @schema
+      bootstrapRootCert: ~
 endpointHealthChecking:
   # -- Enable connectivity health checking between virtual endpoints.
   enabled: true
+# -- Max duration to wait for envoy to respond to configuration changes. Default "10s".
+endpointPolicyUpdateTimeoutDuration: null
 endpointRoutes:
   # @schema
   # type: [boolean, string]
@@ -2339,8 +2441,6 @@ enableMasqueradeRouteSource: false
 enableIPv4BIGTCP: false
 # -- Enables IPv6 BIG TCP support which increases maximum IPv6 GSO/GRO limits for nodes and pods
 enableIPv6BIGTCP: false
-# -- Enable BIG TCP in tunneling mode and increase maximum GRO/GSO limits for VXLAN/GENEVE tunnels
-enableTunnelBIGTCP: false
 
 nat:
   # -- Number of the top-k SNAT map connections to track in Cilium statedb.
@@ -2420,7 +2520,7 @@ loadBalancer:
 
   # -- serviceTopology enables K8s Topology Aware Hints -based service
   # endpoints filtering
-  # serviceTopology: false
+  serviceTopology: false
 
   # -- L7 LoadBalancer
   l7:
diff --git a/packages/system/cilium/images/cilium/Dockerfile b/packages/system/cilium/images/cilium/Dockerfile
index b62f9b46..32fc6fb8 100644
--- a/packages/system/cilium/images/cilium/Dockerfile
+++ b/packages/system/cilium/images/cilium/Dockerfile
@@ -1,2 +1,2 @@
-ARG VERSION=v1.19.1
+ARG VERSION=v1.19.3
 FROM quay.io/cilium/cilium:${VERSION}

From a78505e932650d6dcdab69feb01f6bb3c24f4807 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 13:50:06 +0300
Subject: [PATCH 44/82] chore(cilium): refresh image digest for v1.19.3
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Built ghcr.io/cozystack/cozystack/cilium from the refreshed upstream
v1.19.3 base image and updated values.yaml with the new digest.

Previously values.yaml still pointed at the v1.19.1 cozystack rebuild
by digest while Chart.yaml and the Dockerfile were on v1.19.3 — with
chart default useDigest=true that would have silently pulled v1.19.1
until the next release-tag rebuild.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/system/cilium/values.yaml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/packages/system/cilium/values.yaml b/packages/system/cilium/values.yaml
index cfb2cd6b..c1da6c38 100644
--- a/packages/system/cilium/values.yaml
+++ b/packages/system/cilium/values.yaml
@@ -15,8 +15,8 @@ cilium:
     mode: "kubernetes"
   image:
     repository: ghcr.io/cozystack/cozystack/cilium
-    tag: 1.19.1
-    digest: "sha256:ab3acf270821df4614a8456348a4e0d3098aed72a4b2016a0edfa30d91428c3d"
+    tag: latest
+    digest: "sha256:8f5ab52982fc848ee098ff89919e3528aa8f3a553b82106a25149dcd0b87ea7e"
   envoy:
     enabled: false
   rollOutCiliumPods: true

From 3f36a1b45b2406bc70ab49b4799d45ee88868e48 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 17:14:23 +0300
Subject: [PATCH 45/82] fix(cilium): rebuild image multi-arch and pin tag to
 1.19.3

The previous image digest in values.yaml pointed at a single-arch
linux/arm64 manifest because 'make image' was run from an arm64 host
with the default buildx platform. Cozystack targets amd64 (Talos build
output, E2E runners, most real-world clusters) and also arm64 for
hybrid fleets, so Helm install would fail on amd64 nodes with 'no
matching manifest for linux/amd64 in the manifest list entries'
whenever somebody installed directly from this commit between merge
and the next release-tag CI rebuild.

Fix: rebuilt the image locally with
PLATFORM='linux/amd64,linux/arm64' make image from a buildx
docker-container driver, pushed the multi-arch manifest, and
refreshed values.yaml with:

- digest of the new multi-arch manifest list (verified via
  'docker manifest inspect': amd64 sha256:e1977323..., arm64
  sha256:8f5ab529...).
- tag bumped from 'latest' (emitted by the common-envs.mk settag
  macro on a non-tagged checkout) to '1.19.3', matching the
  established convention in every other packages/system/*/values.yaml
  so reviewers and incident response have a human-readable version
  anchor independent of digest chasing.

The Makefile is left untouched so the CI builder (which only uses the
default docker driver) keeps building single-arch for whatever
architecture it runs on; multi-arch is a responsibility of the
release-tag pipeline or an explicit local rebuild.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/system/cilium/values.yaml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/packages/system/cilium/values.yaml b/packages/system/cilium/values.yaml
index c1da6c38..9aba7cb9 100644
--- a/packages/system/cilium/values.yaml
+++ b/packages/system/cilium/values.yaml
@@ -15,8 +15,8 @@ cilium:
     mode: "kubernetes"
   image:
     repository: ghcr.io/cozystack/cozystack/cilium
-    tag: latest
-    digest: "sha256:8f5ab52982fc848ee098ff89919e3528aa8f3a553b82106a25149dcd0b87ea7e"
+    tag: 1.19.3
+    digest: "sha256:700f06f4803a838a8e830be5ace4650e3ad82bdefabfb2f4d110368d307a5efb"
   envoy:
     enabled: false
   rollOutCiliumPods: true

From adc7abe5c16fc81ce1c063d153d9a920c0aac097 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 14:55:21 +0300
Subject: [PATCH 46/82] feat(ingress): add loadBalancer exposure mode via
 CiliumLoadBalancerIPPool

Service.spec.externalIPs is deprecated upstream in Kubernetes v1.36
(KEP-5707, kubernetes#137293). The AllowServiceExternalIPs feature gate
is expected to default to off around v1.40 and the implementation to
be removed around v1.43. For bare-metal installs that rely on
externalIPs today, cozystack needs a migration path.

This change adds an opt-in 'loadBalancer' exposure mode for the
ingress-nginx Service:

- New platform value 'publishing.exposure' (enum: externalIPs |
  loadBalancer, default externalIPs). Plumbed through cozystack-values
  into each tenant's ingress HelmRelease via the new 'expose-mode' key.
- Unknown values and loadBalancer with an empty externalIPs list fail
  the chart render with explicit error messages, rather than silently
  producing a broken Service.
- When exposure=loadBalancer and the current namespace matches
  publishing.ingressName, the Service becomes type: LoadBalancer with
  externalTrafficPolicy: Local.
- A new template renders a CiliumLoadBalancerIPPool whose blocks come
  from publishing.externalIPs (IPv4 addresses get /32, IPv6 addresses
  get /128) and whose serviceSelector uses Cilium's synthetic
  io.kubernetes.service.namespace key combined with the standard
  app.kubernetes.io/name: ingress-nginx label. No custom label is
  written to the Service itself, avoiding cross-tenant collisions from
  user-defined labels.

Default behaviour is unchanged: without opting in, the Service is
still ClusterIP + spec.externalIPs as today.

Scope: only ingress-nginx is migrated by this setting. Other cozystack
components that still write Service.spec.externalIPs directly (notably
the vpn app) must be migrated separately before the v1.40 feature gate
flip.

Tests: packages/extra/ingress/tests/exposure_test.yaml adds 13
helm-unittest cases covering both modes, IPv4/IPv6, empty-token
filtering, unknown-mode rejection, and the non-matching-namespace
fallback.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/core/platform/templates/apps.yaml    |   1 +
 packages/core/platform/values.yaml            |  35 +++
 packages/extra/ingress/Makefile               |   3 +
 packages/extra/ingress/README.md              |  14 +
 .../ingress/templates/cilium-lb-pool.yaml     |  25 ++
 .../ingress/templates/nginx-ingress.yaml      |  21 +-
 .../extra/ingress/tests/exposure_test.yaml    | 251 ++++++++++++++++++
 7 files changed, 348 insertions(+), 2 deletions(-)
 create mode 100644 packages/extra/ingress/templates/cilium-lb-pool.yaml
 create mode 100644 packages/extra/ingress/tests/exposure_test.yaml

diff --git a/packages/core/platform/templates/apps.yaml b/packages/core/platform/templates/apps.yaml
index 7ee0de52..6e1ae25a 100644
--- a/packages/core/platform/templates/apps.yaml
+++ b/packages/core/platform/templates/apps.yaml
@@ -30,6 +30,7 @@ stringData:
       expose-services: {{ .Values.publishing.exposedServices | join "," | quote }}
       expose-ingress: {{ .Values.publishing.ingressName | quote }}
       expose-external-ips: {{ .Values.publishing.externalIPs | join "," | quote }}
+      expose-mode: {{ .Values.publishing.exposure | default "externalIPs" | quote }}
       cluster-domain: {{ .Values.networking.clusterDomain | quote }}
       api-server-endpoint: {{ .Values.publishing.apiServerEndpoint | quote }}
       {{- with .Values.branding }}
diff --git a/packages/core/platform/values.yaml b/packages/core/platform/values.yaml
index f33926db..605e23bc 100644
--- a/packages/core/platform/values.yaml
+++ b/packages/core/platform/values.yaml
@@ -45,6 +45,41 @@ publishing:
     - cdi-uploadproxy
   apiServerEndpoint: "" # example: "https://api.example.org"
   externalIPs: []
+  # Exposure mode for the ingress-nginx Service. When "externalIPs" (current
+  # default) is selected, the Service is created as ClusterIP with
+  # Service.spec.externalIPs set from publishing.externalIPs. When
+  # "loadBalancer" is selected, the Service is type: LoadBalancer and a
+  # CiliumLoadBalancerIPPool makes those same addresses allocatable via LB IPAM.
+  #
+  # Service.spec.externalIPs is deprecated upstream in Kubernetes v1.36
+  # (KEP-5707). The AllowServiceExternalIPs feature gate is expected to default
+  # to false around v1.40 and the implementation removed around v1.43 — switch
+  # to "loadBalancer" before upgrading past v1.40.
+  #
+  # Caveats for the "loadBalancer" mode:
+  #   - publishing.externalIPs must contain at least one non-empty address,
+  #     otherwise the chart render fails with an explicit error (a LoadBalancer
+  #     Service without a pool would sit in <pending> forever).
+  #   - The ingress-nginx Service is created with externalTrafficPolicy: Local
+  #     to preserve the client source IP. Traffic arriving on a node that does
+  #     not host an ingress-nginx pod is dropped, so the external IP must be
+  #     routed to a node that runs the ingress pod (floating IP / keepalived /
+  #     upstream router / podAntiAffinity).
+  #   - Cilium does NOT announce the IP on its own unless L2 announcements or
+  #     BGP are enabled in the Cilium values (disabled by default in Cozystack).
+  #     This mode assumes the operator already routes the externalIPs to a
+  #     cluster node; enabling announcements is out of scope for this setting.
+  #   - Switching this value on a running cluster causes the ingress-nginx
+  #     Service to be recreated (the HelmRelease has upgrade.force: true and
+  #     the Service kind changes between ClusterIP and LoadBalancer). Expect a
+  #     brief interruption of ingress traffic during the flip.
+  #
+  # Scope: this setting only controls the ingress-nginx Service. Other
+  # cozystack components that currently write Service.spec.externalIPs directly
+  # (e.g. the vpn app at packages/apps/vpn/templates/service.yaml) are NOT
+  # migrated by flipping this value and must be addressed separately before
+  # the AllowServiceExternalIPs feature gate flips to off in ~v1.40.
+  exposure: externalIPs # "externalIPs" or "loadBalancer"
   certificates:
     solver: http01 # "http01" or "dns01"
     issuerName: letsencrypt-prod
diff --git a/packages/extra/ingress/Makefile b/packages/extra/ingress/Makefile
index 958ce484..65b53c03 100644
--- a/packages/extra/ingress/Makefile
+++ b/packages/extra/ingress/Makefile
@@ -10,3 +10,6 @@ get-cloudflare-ips:
 generate:
 	cozyvalues-gen -v values.yaml -s values.schema.json -r README.md
 	../../../hack/update-crd.sh
+
+test:
+	helm unittest .
diff --git a/packages/extra/ingress/README.md b/packages/extra/ingress/README.md
index 0e786dfb..c541d8e9 100644
--- a/packages/extra/ingress/README.md
+++ b/packages/extra/ingress/README.md
@@ -14,3 +14,17 @@
 | `resources.memory` | Memory (RAM) available to each replica.                                                                                                 | `quantity` | `""`    |
 | `resourcesPreset`  | Default sizing preset used when `resources` is omitted.                                                                                 | `string`   | `micro` |
 
+
+## Exposure mode
+
+The ingress Service type is driven by the cluster-wide `publishing.exposure` value in the platform chart, not by any key in this package. Two modes exist:
+
+- `externalIPs` (default) has three rendered shapes:
+  - Release namespace matches `publishing.ingressName` AND `publishing.externalIPs` is non-empty → Service is `ClusterIP` with `Service.spec.externalIPs` set from that list and `externalTrafficPolicy: Cluster`.
+  - Release namespace matches `publishing.ingressName` but `publishing.externalIPs` is empty → Service falls back to `type: LoadBalancer` with `externalTrafficPolicy: Local`.
+  - Release namespace does not match `publishing.ingressName` (non-root tenants) → Service is `type: LoadBalancer` with `externalTrafficPolicy: Local`.
+  `Service.spec.externalIPs` is deprecated upstream in Kubernetes v1.36 (KEP-5707); plan migration before v1.40.
+- `loadBalancer` — Service is `type: LoadBalancer` with `externalTrafficPolicy: Local`, and a `CiliumLoadBalancerIPPool` makes the addresses in `publishing.externalIPs` allocatable via Cilium LB IPAM. Requires `publishing.externalIPs` to contain at least one non-empty address (render fails otherwise) and assumes the addresses are already routed to a cluster node (floating IP / upstream router). See the inline comment on `publishing.exposure` in the platform chart for full caveats, including the note that switching the value on a running cluster causes the ingress Service to be recreated.
+
+This setting only migrates ingress-nginx away from `Service.spec.externalIPs`. Other cozystack components that use the same deprecated field (e.g. the `vpn` app) must be migrated separately before Kubernetes v1.40 flips the `AllowServiceExternalIPs` feature gate off.
+
diff --git a/packages/extra/ingress/templates/cilium-lb-pool.yaml b/packages/extra/ingress/templates/cilium-lb-pool.yaml
new file mode 100644
index 00000000..a383922b
--- /dev/null
+++ b/packages/extra/ingress/templates/cilium-lb-pool.yaml
@@ -0,0 +1,25 @@
+{{- $exposeIngress := (index .Values._cluster "expose-ingress") | default "tenant-root" }}
+{{- $exposeMode := (index .Values._cluster "expose-mode") | default "externalIPs" }}
+{{- $exposeExternalIPs := (index .Values._cluster "expose-external-ips") | default "" | nospace }}
+{{- $exposeIPsList := list }}
+{{- range splitList "," $exposeExternalIPs }}
+  {{- $ip := . | trim }}
+  {{- if $ip }}
+    {{- $exposeIPsList = append $exposeIPsList $ip }}
+  {{- end }}
+{{- end }}
+{{- if and (eq $exposeMode "loadBalancer") (eq $exposeIngress .Release.Namespace) $exposeIPsList }}
+apiVersion: cilium.io/v2
+kind: CiliumLoadBalancerIPPool
+metadata:
+  name: {{ trimPrefix "tenant-" .Release.Namespace }}-ingress
+spec:
+  blocks:
+  {{- range $exposeIPsList }}
+  - cidr: {{ . }}/{{ if contains ":" . }}128{{ else }}32{{ end }}
+  {{- end }}
+  serviceSelector:
+    matchLabels:
+      "io.kubernetes.service.namespace": {{ .Release.Namespace | quote }}
+      "app.kubernetes.io/name": ingress-nginx
+{{- end }}
diff --git a/packages/extra/ingress/templates/nginx-ingress.yaml b/packages/extra/ingress/templates/nginx-ingress.yaml
index ca50d276..8e1f00fb 100644
--- a/packages/extra/ingress/templates/nginx-ingress.yaml
+++ b/packages/extra/ingress/templates/nginx-ingress.yaml
@@ -1,5 +1,19 @@
 {{- $exposeIngress := (index .Values._cluster "expose-ingress") | default "tenant-root" }}
 {{- $exposeExternalIPs := (index .Values._cluster "expose-external-ips") | default "" | nospace }}
+{{- $exposeMode := (index .Values._cluster "expose-mode") | default "externalIPs" }}
+{{- $exposeIPsList := list }}
+{{- range splitList "," $exposeExternalIPs }}
+  {{- $ip := . | trim }}
+  {{- if $ip }}
+    {{- $exposeIPsList = append $exposeIPsList $ip }}
+  {{- end }}
+{{- end }}
+{{- if not (has $exposeMode (list "externalIPs" "loadBalancer")) }}
+{{- fail (printf "unknown publishing.exposure mode %q: must be \"externalIPs\" or \"loadBalancer\"" $exposeMode) }}
+{{- end }}
+{{- if and (eq $exposeMode "loadBalancer") (eq $exposeIngress .Release.Namespace) (not $exposeIPsList) }}
+{{- fail "publishing.exposure=loadBalancer requires publishing.externalIPs to contain at least one non-empty address: the CiliumLoadBalancerIPPool has nothing to advertise and the Service would stay in <pending>." }}
+{{- end }}
 apiVersion: helm.toolkit.fluxcd.io/v2
 kind: HelmRelease
 metadata:
@@ -41,9 +55,12 @@ spec:
           enabled: false
         {{- end }}
         service:
-          {{- if and (eq $exposeIngress .Release.Namespace) $exposeExternalIPs }}
+          {{- if and (eq $exposeIngress .Release.Namespace) (eq $exposeMode "loadBalancer") }}
+          type: LoadBalancer
+          externalTrafficPolicy: Local
+          {{- else if and (eq $exposeIngress .Release.Namespace) $exposeIPsList }}
           externalIPs:
-            {{- toYaml (splitList "," $exposeExternalIPs) | nindent 12 }}
+            {{- toYaml $exposeIPsList | nindent 12 }}
           type: ClusterIP
           externalTrafficPolicy: Cluster
           {{- else }}
diff --git a/packages/extra/ingress/tests/exposure_test.yaml b/packages/extra/ingress/tests/exposure_test.yaml
new file mode 100644
index 00000000..780e0715
--- /dev/null
+++ b/packages/extra/ingress/tests/exposure_test.yaml
@@ -0,0 +1,251 @@
+suite: ingress exposure modes
+templates:
+  - templates/nginx-ingress.yaml
+  - templates/cilium-lb-pool.yaml
+
+release:
+  name: ingress
+  namespace: tenant-root
+
+tests:
+  - it: default exposure (externalIPs) renders ClusterIP Service with spec.externalIPs and no CiliumLoadBalancerIPPool
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,192.0.2.11"
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.type
+          value: ClusterIP
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalTrafficPolicy
+          value: Cluster
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalIPs
+          value:
+            - 192.0.2.10
+            - 192.0.2.11
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.labels
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 0
+
+  - it: legacy config without expose-mode falls back to externalIPs behavior
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.type
+          value: ClusterIP
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 0
+
+  - it: externalIPs mode in a namespace other than publishing.ingressName renders LoadBalancer fallback without externalIPs
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+    release:
+      namespace: tenant-other
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.type
+          value: LoadBalancer
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalTrafficPolicy
+          value: Local
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.externalIPs
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 0
+
+  - it: loadBalancer mode renders LoadBalancer Service with lb-pool label and a v2 CiliumLoadBalancerIPPool
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,192.0.2.11"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.type
+          value: LoadBalancer
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalTrafficPolicy
+          value: Local
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.labels
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.externalIPs
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 1
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: apiVersion
+          value: cilium.io/v2
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: kind
+          value: CiliumLoadBalancerIPPool
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: metadata.name
+          value: root-ingress
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 192.0.2.11/32
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.serviceSelector.matchLabels["io.kubernetes.service.namespace"]
+          value: tenant-root
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.serviceSelector.matchLabels["app.kubernetes.io/name"]
+          value: ingress-nginx
+
+  - it: loadBalancer mode with IPv6 address emits /128 CIDR
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "2001:db8::1"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.blocks
+          value:
+            - cidr: 2001:db8::1/128
+
+  - it: loadBalancer mode with mixed IPv4 and IPv6 emits correct CIDR per family
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,2001:db8::1"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 2001:db8::1/128
+
+  - it: loadBalancer mode without externalIPs fails chart render with explicit message
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: ""
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        failedTemplate:
+          errorMessage: "publishing.exposure=loadBalancer requires publishing.externalIPs to contain at least one non-empty address: the CiliumLoadBalancerIPPool has nothing to advertise and the Service would stay in <pending>."
+
+  - it: unknown exposure mode is rejected with a clear error (case-sensitive enum)
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+        expose-mode: LoadBalancer
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        failedTemplate:
+          errorMessage: "unknown publishing.exposure mode \"LoadBalancer\": must be \"externalIPs\" or \"loadBalancer\""
+
+  - it: another typo in exposure mode also fails
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+        expose-mode: loadbalancer
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        failedTemplate:
+          errorMessage: "unknown publishing.exposure mode \"loadbalancer\": must be \"externalIPs\" or \"loadBalancer\""
+
+  - it: loadBalancer mode in a namespace other than publishing.ingressName falls back to LoadBalancer Service without pool
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+        expose-mode: loadBalancer
+    release:
+      namespace: tenant-other
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.type
+          value: LoadBalancer
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalTrafficPolicy
+          value: Local
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.labels
+      - template: templates/nginx-ingress.yaml
+        notExists:
+          path: spec.values.ingress-nginx.controller.service.externalIPs
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 0
+
+  - it: loadBalancer mode filters out empty entries from externalIPs (trailing comma, leading comma)
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,,192.0.2.11,"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/cilium-lb-pool.yaml
+        hasDocuments:
+          count: 1
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 192.0.2.11/32
+
+  - it: loadBalancer mode with only-empty externalIPs fails chart render (comma-only input)
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: ",,"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        failedTemplate:
+          errorMessage: "publishing.exposure=loadBalancer requires publishing.externalIPs to contain at least one non-empty address: the CiliumLoadBalancerIPPool has nothing to advertise and the Service would stay in <pending>."
+
+  - it: externalIPs mode also filters out empty entries (trailing comma)
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,"
+    asserts:
+      - template: templates/nginx-ingress.yaml
+        equal:
+          path: spec.values.ingress-nginx.controller.service.externalIPs
+          value:
+            - 192.0.2.10

From c34a9db6bd2d052e290cb550713369e82d1f3d0a Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 17:59:49 +0300
Subject: [PATCH 47/82] docs(agents): broaden make generate example to apps
 packages

Address review feedback from coderabbitai on docs/agents/contributing.md:26:
Replace the hard-coded packages/extra/<name> path in the example with
packages/<apps-or-extra>/<name> so the example matches the preceding
text that describes both apps and extra packages.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 9626e81e..5b1c6ba5 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -21,8 +21,8 @@ Several files in each package are produced by `make generate` from `values.yaml`
 **Before committing edits to any of those sources**, run `make generate` inside the package and stage the full diff:
 
 ```bash
-make -C packages/extra/<name> generate
-git add packages/extra/<name>/ packages/system/<name>-rd/
+make -C packages/<apps-or-extra>/<name> generate
+git add packages/<apps-or-extra>/<name>/ packages/system/<name>-rd/
 ```
 
 The repo's pre-commit CI job runs `make generate` in every package and then `git diff --exit-code`. Any unstaged generator output fails the job with exit code 123 and blocks the PR. Also rerun `make generate` after a `git commit --amend` if the amended change touched any of the sources above.

From 41fd80711bbe4a3cdb12b3af0bcd8bfdd40993e5 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 18:00:07 +0300
Subject: [PATCH 48/82] docs(agents): fix grammar in regen discovery hint

Address review feedback from gemini-code-assist on docs/agents/contributing.md:30:
Reword "likely to need regenerated" (regional construction) to
"likely needs to be regenerated" for standard technical prose.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 5b1c6ba5..cef12494 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -27,7 +27,7 @@ git add packages/<apps-or-extra>/<name>/ packages/system/<name>-rd/
 
 The repo's pre-commit CI job runs `make generate` in every package and then `git diff --exit-code`. Any unstaged generator output fails the job with exit code 123 and blocks the PR. Also rerun `make generate` after a `git commit --amend` if the amended change touched any of the sources above.
 
-To locate packages a WIP branch is likely to need regenerated:
+To locate packages a WIP branch likely needs to be regenerated:
 
 ```bash
 git diff --name-only | xargs -n1 dirname | sort -u | grep ^packages/

From 4813566a30fb60eca00ce0ac83254daabf65341f Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 18:02:04 +0300
Subject: [PATCH 49/82] docs(agents): mark scopes list as illustrative examples

Address review feedback from gemini-code-assist on docs/agents/contributing.md:11:
Scope linters kept flagging valid scopes like 'agents' as unknown because
the list read as exhaustive. Annotate it as examples (not exhaustive) and
add 'agents' to the Other group so both humans and review bots stop
tripping on scopes that are already in regular use across the repo
history.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index cef12494..658dab74 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -43,10 +43,10 @@ git commit --signoff -m "type(scope): brief description"
 
 **Types:** `feat`, `fix`, `docs`, `style`, `refactor`, `perf`, `test`, `build`, `ci`, `chore`
 
-**Scopes:**
+**Scopes** (e.g., not exhaustive — use any scope that names the component you are touching):
 - System: `dashboard`, `platform`, `cilium`, `kube-ovn`, `linstor`, `fluxcd`, `cluster-api`
 - Apps: `postgres`, `mariadb`, `redis`, `kafka`, `clickhouse`, `virtual-machine`, `kubernetes`
-- Other: `api`, `hack`, `tests`, `ci`, `docs`, `maintenance`
+- Other: `api`, `hack`, `tests`, `ci`, `docs`, `agents`, `maintenance`
 
 Breaking changes: append `!` after type/scope (`feat(api)!: ...`) or add a `BREAKING CHANGE:` footer.
 

From 64a3edff01533a0483db31686a6385dffa81ee55 Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Thu, 23 Apr 2026 18:13:10 +0300
Subject: [PATCH 50/82] fix(api): prevent IDOR in TenantNamespace Get and Watch
 handlers

Fixed two IDOR vulnerabilities allowing authenticated users to access
metadata of any tenant namespace without proper authorization.

Changes:
- Added hasAccessToNamespace() for efficient single-namespace access checks
- Get() now verifies access before returning namespace metadata
- Watch() filters events per-namespace with proper authorization
- Returns NotFound (not Forbidden) to prevent tenant enumeration

Performance optimization:
- hasAccessToNamespace() lists RoleBindings only in target namespace
  instead of listing all cluster RoleBindings (order of magnitude faster)
- Watch handler logs authorization errors for security audit

Additional fixes:
- Handle ServiceAccount subjects with empty namespace correctly
- Add klog error logging for failed authorization checks

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 pkg/registry/core/tenantnamespace/rest.go | 89 ++++++++++++++++++++++-
 1 file changed, 87 insertions(+), 2 deletions(-)

diff --git a/pkg/registry/core/tenantnamespace/rest.go b/pkg/registry/core/tenantnamespace/rest.go
index f1ed3fab..7724c0e1 100644
--- a/pkg/registry/core/tenantnamespace/rest.go
+++ b/pkg/registry/core/tenantnamespace/rest.go
@@ -24,6 +24,7 @@ import (
 	"k8s.io/apimachinery/pkg/watch"
 	"k8s.io/apiserver/pkg/endpoints/request"
 	"k8s.io/apiserver/pkg/registry/rest"
+	"k8s.io/klog/v2"
 	"sigs.k8s.io/controller-runtime/pkg/client"
 
 	corev1alpha1 "github.com/cozystack/cozystack/pkg/apis/core/v1alpha1"
@@ -123,8 +124,18 @@ func (r *REST) Get(
 		return nil, apierrors.NewNotFound(r.gvr.GroupResource(), name)
 	}
 
+	// Check if user has access to this namespace
+	hasAccess, err := r.hasAccessToNamespace(ctx, name)
+	if err != nil {
+		return nil, err
+	}
+	if !hasAccess {
+		// Return NotFound instead of Forbidden to prevent enumeration
+		return nil, apierrors.NewNotFound(r.gvr.GroupResource(), name)
+	}
+
 	ns := &corev1.Namespace{}
-	err := r.c.Get(ctx, types.NamespacedName{Namespace: "", Name: name}, ns, &client.GetOptions{Raw: opts})
+	err = r.c.Get(ctx, types.NamespacedName{Namespace: "", Name: name}, ns, &client.GetOptions{Raw: opts})
 	if err != nil {
 		return nil, err
 	}
@@ -189,6 +200,17 @@ func (r *REST) Watch(ctx context.Context, opts *metainternal.ListOptions) (watch
 				continue
 			}
 
+			// Check if user has access to this namespace
+			hasAccess, err := r.hasAccessToNamespace(ctx, ns.Name)
+			if err != nil {
+				klog.Errorf("Failed to check access for namespace %s in watch: %v", ns.Name, err)
+				continue
+			}
+			if !hasAccess {
+				// User doesn't have access, skip this event
+				continue
+			}
+
 			out := &corev1alpha1.TenantNamespace{
 				TypeMeta: metav1.TypeMeta{
 					APIVersion: corev1alpha1.SchemeGroupVersion.String(),
@@ -359,7 +381,11 @@ func (r *REST) filterAccessible(
 					break subjectLoop
 				}
 			case "ServiceAccount":
-				if u.GetName() == fmt.Sprintf("system:serviceaccount:%s:%s", subj.Namespace, subj.Name) {
+				saNamespace := subj.Namespace
+				if saNamespace == "" {
+					saNamespace = rbs.Items[i].Namespace
+				}
+				if u.GetName() == fmt.Sprintf("system:serviceaccount:%s:%s", saNamespace, subj.Name) {
 					allowedNameSet[rbs.Items[i].Namespace] = struct{}{}
 					break subjectLoop
 				}
@@ -373,6 +399,65 @@ func (r *REST) filterAccessible(
 	return allowed, nil
 }
 
+// hasAccessToNamespace checks if the user has access to a single namespace.
+// This is optimized for Get/Watch operations where we check one namespace at a time.
+// It lists RoleBindings only in the target namespace instead of all cluster RoleBindings.
+func (r *REST) hasAccessToNamespace(
+	ctx context.Context,
+	namespace string,
+) (bool, error) {
+	u, ok := request.UserFrom(ctx)
+	if !ok {
+		return false, fmt.Errorf("user missing in context")
+	}
+
+	// Check privileged groups
+	groups := make(map[string]struct{})
+	for _, group := range u.GetGroups() {
+		groups[group] = struct{}{}
+	}
+	if _, ok := groups["system:masters"]; ok {
+		return true, nil
+	}
+	if _, ok := groups["cozystack-cluster-admin"]; ok {
+		return true, nil
+	}
+
+	// List RoleBindings only in the target namespace
+	rbs := &rbacv1.RoleBindingList{}
+	err := r.c.List(ctx, rbs, client.InNamespace(namespace))
+	if err != nil {
+		return false, fmt.Errorf("failed to list rolebindings in %s: %w", namespace, err)
+	}
+
+	// Check if user is in any RoleBinding subjects
+	for i := range rbs.Items {
+		for j := range rbs.Items[i].Subjects {
+			subj := rbs.Items[i].Subjects[j]
+			switch subj.Kind {
+			case "Group":
+				if _, ok := groups[subj.Name]; ok {
+					return true, nil
+				}
+			case "User":
+				if subj.Name == u.GetName() {
+					return true, nil
+				}
+			case "ServiceAccount":
+				saNamespace := subj.Namespace
+				if saNamespace == "" {
+					saNamespace = rbs.Items[i].Namespace
+				}
+				if u.GetName() == fmt.Sprintf("system:serviceaccount:%s:%s", saNamespace, subj.Name) {
+					return true, nil
+				}
+			}
+		}
+	}
+
+	return false, nil
+}
+
 // -----------------------------------------------------------------------------
 // Boiler-plate
 // -----------------------------------------------------------------------------

From 5b2501db91c82d8f9ce1d7187325927f7504847c Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Thu, 23 Apr 2026 18:17:47 +0300
Subject: [PATCH 51/82] test(api): add security tests for TenantNamespace IDOR
 fix

Added comprehensive unit tests for authorization logic:

- hasAccessToNamespace() tests:
  - User subject access (positive and negative cases)
  - Group subject access
  - ServiceAccount subject access
  - ServiceAccount with empty namespace (defaults to RoleBinding ns)
  - Privileged groups (system:masters, cozystack-cluster-admin)

- Get() handler tests:
  - Returns namespace when user has access
  - Returns NotFound when user lacks access (not Forbidden)
  - Returns NotFound for non-tenant namespaces

All tests verify that authorization correctly enforces RoleBinding-based
access control and prevents IDOR vulnerabilities.

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 .../core/tenantnamespace/rest_test.go         | 461 ++++++++++++++++++
 1 file changed, 461 insertions(+)

diff --git a/pkg/registry/core/tenantnamespace/rest_test.go b/pkg/registry/core/tenantnamespace/rest_test.go
index 7f2979bc..eb678949 100644
--- a/pkg/registry/core/tenantnamespace/rest_test.go
+++ b/pkg/registry/core/tenantnamespace/rest_test.go
@@ -3,10 +3,18 @@
 package tenantnamespace
 
 import (
+	"context"
 	"testing"
 
 	corev1 "k8s.io/api/core/v1"
+	rbacv1 "k8s.io/api/rbac/v1"
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/runtime"
+	"k8s.io/apimachinery/pkg/runtime/schema"
+	"k8s.io/apiserver/pkg/authentication/user"
+	"k8s.io/apiserver/pkg/endpoints/request"
+	"sigs.k8s.io/controller-runtime/pkg/client/fake"
 )
 
 func TestMakeListSortsAlphabetically(t *testing.T) {
@@ -38,3 +46,456 @@ func TestMakeListSortsAlphabetically(t *testing.T) {
 		}
 	}
 }
+
+// Security tests for IDOR fix
+
+func TestHasAccessToNamespace_WithUserAccess(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "User", Name: "test-user", APIGroup: "rbac.authorization.k8s.io"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected user to have access, but got false")
+	}
+}
+
+func TestHasAccessToNamespace_WithoutAccess(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	// RoleBinding for different user
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "User", Name: "other-user", APIGroup: "rbac.authorization.k8s.io"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if hasAccess {
+		t.Error("expected user to NOT have access, but got true")
+	}
+}
+
+func TestHasAccessToNamespace_WithGroupAccess(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "Group", Name: "test-group", APIGroup: "rbac.authorization.k8s.io"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:authenticated", "test-group"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected user to have access via group, but got false")
+	}
+}
+
+func TestHasAccessToNamespace_SystemMasters(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "admin",
+		Groups: []string{"system:masters"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected system:masters to have access, but got false")
+	}
+}
+
+func TestHasAccessToNamespace_CozyAdminGroup(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "cozy-admin",
+		Groups: []string{"cozystack-cluster-admin"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected cozystack-cluster-admin to have access, but got false")
+	}
+}
+
+func TestHasAccessToNamespace_ServiceAccount(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "ServiceAccount", Name: "test-sa", Namespace: "tenant-test"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "system:serviceaccount:tenant-test:test-sa",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected service account to have access, but got false")
+	}
+}
+
+func TestHasAccessToNamespace_ServiceAccountEmptyNamespace(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	// ServiceAccount subject with empty namespace should default to RoleBinding namespace
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "ServiceAccount", Name: "test-sa", Namespace: ""}, // Empty namespace
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "system:serviceaccount:tenant-test:test-sa",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	hasAccess, err := r.hasAccessToNamespace(ctx, "tenant-test")
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if !hasAccess {
+		t.Error("expected service account with empty namespace to have access, but got false")
+	}
+}
+
+func TestGet_WithAccess(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "User", Name: "test-user", APIGroup: "rbac.authorization.k8s.io"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	obj, err := r.Get(ctx, "tenant-test", &metav1.GetOptions{})
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if obj == nil {
+		t.Fatal("expected object, got nil")
+	}
+}
+
+func TestGet_WithoutAccess(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	ns := &corev1.Namespace{
+		ObjectMeta: metav1.ObjectMeta{Name: "tenant-test"},
+	}
+
+	// RoleBinding for different user
+	rb := &rbacv1.RoleBinding{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "test-binding",
+			Namespace: "tenant-test",
+		},
+		Subjects: []rbacv1.Subject{
+			{Kind: "User", Name: "other-user", APIGroup: "rbac.authorization.k8s.io"},
+		},
+		RoleRef: rbacv1.RoleRef{
+			APIGroup: "rbac.authorization.k8s.io",
+			Kind:     "Role",
+			Name:     "test-role",
+		},
+	}
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithObjects(ns, rb).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:authenticated"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	obj, err := r.Get(ctx, "tenant-test", &metav1.GetOptions{})
+	if err == nil {
+		t.Fatal("expected error, got nil")
+	}
+	if obj != nil {
+		t.Errorf("expected nil object, got %v", obj)
+	}
+
+	// Verify it returns NotFound (not Forbidden) to prevent enumeration
+	if !apierrors.IsNotFound(err) {
+		t.Errorf("expected NotFound error, got %v", err)
+	}
+}
+
+func TestGet_NonTenantNamespace(t *testing.T) {
+	scheme := runtime.NewScheme()
+	_ = corev1.AddToScheme(scheme)
+	_ = rbacv1.AddToScheme(scheme)
+
+	client := fake.NewClientBuilder().
+		WithScheme(scheme).
+		Build()
+
+	r := &REST{
+		c:   client,
+		gvr: schema.GroupVersionResource{Group: "core.cozystack.io", Version: "v1alpha1", Resource: "tenantnamespaces"},
+	}
+
+	u := &user.DefaultInfo{
+		Name:   "test-user",
+		Groups: []string{"system:masters"},
+	}
+	ctx := request.WithUser(context.Background(), u)
+
+	obj, err := r.Get(ctx, "default", &metav1.GetOptions{})
+	if err == nil {
+		t.Fatal("expected error for non-tenant namespace, got nil")
+	}
+	if obj != nil {
+		t.Errorf("expected nil object, got %v", obj)
+	}
+	if !apierrors.IsNotFound(err) {
+		t.Errorf("expected NotFound error, got %v", err)
+	}
+}

From 9d552d4086726d2f5c6de5adaa2783082893a8c5 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Thu, 23 Apr 2026 22:19:35 +0500
Subject: [PATCH 52/82] ci(api): broaden codegen drift trigger paths and detect
 untracked files

- Add generated-output dirs (pkg/generated, internal/crdinstall/manifests,
  packages/system/*/definitions) and Makefile to the workflow paths: filter
  so PRs that modify only generated artifacts still trigger the drift check.
- Mirror the same paths in the root pre-commit hook's files: regex so
  manual edits to generated files or changes to the root generate target
  re-run make generate through pre-commit.
- Switch drift detection in the workflow from `git diff --exit-code` to
  `git status --porcelain` so new untracked files produced by make generate
  (e.g. generated YAML/Go for a new API type) also fail the job; dump
  `git diff --color=always` on failure for easier debugging.

Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 .github/workflows/codegen-drift.yml | 10 +++++++++-
 .pre-commit-config.yaml             |  2 +-
 2 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/codegen-drift.yml b/.github/workflows/codegen-drift.yml
index b08f5d7e..a26caf5e 100644
--- a/.github/workflows/codegen-drift.yml
+++ b/.github/workflows/codegen-drift.yml
@@ -6,8 +6,15 @@ on:
     paths:
       - 'api/**'
       - 'pkg/apis/**'
+      - 'pkg/generated/**'
+      - 'internal/crdinstall/manifests/**'
+      - 'packages/system/cozystack-controller/definitions/**'
+      - 'packages/system/application-definition-crd/definition/**'
+      - 'packages/system/backup-controller/definitions/**'
+      - 'packages/system/backupstrategy-controller/definitions/**'
       - 'hack/update-codegen.sh'
       - 'hack/boilerplate.go.txt'
+      - 'Makefile'
       - 'go.mod'
       - 'go.sum'
       - '.github/workflows/codegen-drift.yml'
@@ -46,8 +53,9 @@ jobs:
 
       - name: Fail on drift
         run: |
-          if ! git diff --exit-code; then
+          if [ -n "$(git status --porcelain)" ]; then
             echo "::error::'make generate' produced changes. Run 'make generate' locally and commit the result."
             git status --short
+            git diff --color=always
             exit 1
           fi
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 836f79c5..6492b92d 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -10,7 +10,7 @@ repos:
           git diff --color=always | cat
         '
       language: system
-      files: ^(api/|pkg/apis/|hack/update-codegen\.sh$|hack/boilerplate\.go\.txt$)
+      files: ^(api/|pkg/apis/|pkg/generated/|internal/crdinstall/manifests/|packages/system/cozystack-controller/definitions/|packages/system/application-definition-crd/definition/|packages/system/backup-controller/definitions/|packages/system/backupstrategy-controller/definitions/|hack/update-codegen\.sh$|hack/boilerplate\.go\.txt$|Makefile$)
       pass_filenames: false
     - id: run-make-generate
       name: Run 'make generate' in all app directories

From 6af74cec0ea436cbd5aefce5e5c8bedf86e5b81a Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Thu, 23 Apr 2026 23:24:38 +0300
Subject: [PATCH 53/82] fix(ingress): accept pre-CIDR externalIPs in
 CiliumLoadBalancerIPPool

Address review feedback from gemini-code-assist on
packages/extra/ingress/templates/cilium-lb-pool.yaml:19: if the operator
passes an externalIP already in CIDR form (192.0.2.10/32 or 2001:db8::1/128),
the template appended a second /32 or /128 suffix producing an invalid
CiliumLoadBalancerIPPool block. Guard the suffix append on the absence of
"/" in the input.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../extra/ingress/templates/cilium-lb-pool.yaml    |  2 +-
 packages/extra/ingress/tests/exposure_test.yaml    | 14 ++++++++++++++
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/packages/extra/ingress/templates/cilium-lb-pool.yaml b/packages/extra/ingress/templates/cilium-lb-pool.yaml
index a383922b..048eb58b 100644
--- a/packages/extra/ingress/templates/cilium-lb-pool.yaml
+++ b/packages/extra/ingress/templates/cilium-lb-pool.yaml
@@ -16,7 +16,7 @@ metadata:
 spec:
   blocks:
   {{- range $exposeIPsList }}
-  - cidr: {{ . }}/{{ if contains ":" . }}128{{ else }}32{{ end }}
+  - cidr: {{ . }}{{ if not (contains "/" .) }}/{{ if contains ":" . }}128{{ else }}32{{ end }}{{ end }}
   {{- end }}
   serviceSelector:
     matchLabels:
diff --git a/packages/extra/ingress/tests/exposure_test.yaml b/packages/extra/ingress/tests/exposure_test.yaml
index 780e0715..323c8e68 100644
--- a/packages/extra/ingress/tests/exposure_test.yaml
+++ b/packages/extra/ingress/tests/exposure_test.yaml
@@ -249,3 +249,17 @@ tests:
           path: spec.values.ingress-nginx.controller.service.externalIPs
           value:
             - 192.0.2.10
+
+  - it: loadBalancer mode accepts pre-CIDR input without double-suffixing
+    set:
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10/32,2001:db8::1/128"
+        expose-mode: loadBalancer
+    asserts:
+      - template: templates/cilium-lb-pool.yaml
+        equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 2001:db8::1/128

From 9b87bda06af20d7c1ca8d96b8f338f3bc937d0c5 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Fri, 24 Apr 2026 03:14:41 +0300
Subject: [PATCH 54/82] fix(postgres-operator): block HelmRelease Ready until
 the cnpg webhook actually serves
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The cnpg admission webhook's controller pod passes readinessProbe as soon
as the local HTTPS server binds to :9443. The HelmRelease marks itself
Ready right after that via helm install --wait. But the Service
cnpg-webhook-service needs its EndpointSlice populated and the data plane
(kube-proxy / cilium) programmed before kube-apiserver can reach the
webhook through the Service ClusterIP. That gap is short but not zero,
and any HelmRelease that depends on postgres-operator (cozy-keycloak,
tenant Postgres apps) can fire its own install inside the window and hit

  Internal error occurred: failed calling webhook "mcluster.cnpg.io":
  failed to call webhook: Post "https://cnpg-webhook-service.cozy-postgres-operator.svc:443/...":
  dial tcp <svc-ip>:443: connect: connection refused

which fails the install of the downstream release. Seen on cozystack/cozystack#2470
E2E run 24862782568.

Add a post-install,post-upgrade Helm hook Job that blocks the release
from reporting Ready until the webhook answers /readyz through the
apiserver service proxy. Apiserver proxy routes the call over the same
Service IP → EndpointSlice → pod path the admission webhook uses, so
once it responds, the webhook admission path is also working.

RBAC is minimal: a dedicated ServiceAccount with a ClusterRole that only
grants get on services/proxy scoped to https:cnpg-webhook-service:webhook-server.
The Job times out after 120s with 60 attempts at 2s intervals — longer
than any data-plane programming delay seen on E2E, but bounded.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/system/postgres-operator/Makefile    |   3 +
 .../templates/webhook-ready-hook.yaml         | 139 ++++++++
 .../tests/webhook-ready-hook_test.yaml        | 308 ++++++++++++++++++
 packages/system/postgres-operator/values.yaml |  25 ++
 4 files changed, 475 insertions(+)
 create mode 100644 packages/system/postgres-operator/templates/webhook-ready-hook.yaml
 create mode 100644 packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml

diff --git a/packages/system/postgres-operator/Makefile b/packages/system/postgres-operator/Makefile
index f6d7ddaa..5279f8a7 100644
--- a/packages/system/postgres-operator/Makefile
+++ b/packages/system/postgres-operator/Makefile
@@ -3,6 +3,9 @@ export NAMESPACE=cozy-$(NAME)
 
 include ../../../hack/package.mk
 
+test:
+	helm unittest .
+
 update:
 	rm -rf charts
 	helm repo add cnpg https://cloudnative-pg.github.io/charts
diff --git a/packages/system/postgres-operator/templates/webhook-ready-hook.yaml b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
new file mode 100644
index 00000000..2ad4e7c9
--- /dev/null
+++ b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
@@ -0,0 +1,139 @@
+{{- /*
+  Post-install gate: block the HelmRelease from reporting Ready until the
+  cnpg admission webhook actually serves through the cluster Service. Helm
+  --wait on the controller pod passes once its readinessProbe passes, but
+  EndpointSlice propagation and kube-proxy/cilium data-plane programming
+  can lag by a second or two — long enough for any HelmRelease that
+  depends on postgres-operator (e.g. cozy-keycloak, tenant Postgres apps)
+  to fire its own install and have kube-apiserver hit the mcluster.cnpg.io
+  mutating webhook with "dial tcp <svc-ip>:443: connect: connection refused".
+
+  The Job uses the apiserver service proxy, which exercises the same
+  endpoint-resolution and apiserver-initiated pod dial that the admission
+  webhook path uses. Once /readyz answers through the proxy the data-plane
+  race is resolved. It does not verify the webhook's TLS CA bundle, so
+  this gate is scoped to reachability regressions, not cert rotation.
+
+  The service name and port name are hardcoded literals. Upstream cnpg
+  pins the service name in charts/cloudnative-pg/values.yaml with a
+  comment "DO NOT CHANGE THE SERVICE NAME as it is currently used to
+  generate the certificate and can not be configured". The port name is
+  fixed in charts/cloudnative-pg/templates/service.yaml (ports[0].name:
+  webhook-server). If a future `make update` ever changes either literal
+  upstream, the sync-check helm-unittest test
+  (tests/webhook-ready-hook_test.yaml) renders the subchart Service and
+  fails if the literal drifts — forcing this template to be updated in
+  the same change.
+*/}}
+{{- $_ := required "webhookReady.image.repository must be set to the container image providing kubectl for the post-install readiness Job" .Values.webhookReady.image.repository -}}
+{{- $_ := required "webhookReady.image.tag must be set for the post-install readiness Job" .Values.webhookReady.image.tag -}}
+{{- /* $svcName and $portName are hardcoded literals; see header comment. */ -}}
+{{- $svcName := "cnpg-webhook-service" -}}
+{{- /* $portName is the service port NAME, not number — matches ports[0].name in the vendored subchart's Service. */ -}}
+{{- $portName := "webhook-server" -}}
+{{- $resourceName := printf "https:%s:%s" $svcName $portName -}}
+{{- $maxAttempts := .Values.webhookReady.maxAttempts | default 60 -}}
+{{- $sleepSeconds := .Values.webhookReady.sleepSeconds | default 2 -}}
+{{- /* Derive activeDeadlineSeconds from retries + 60s slack so a values override that raises maxAttempts doesn't get silently cut. */ -}}
+{{- $deadline := add (mul (int $maxAttempts) (int $sleepSeconds)) 60 -}}
+{{- $image := printf "%s:%s" .Values.webhookReady.image.repository .Values.webhookReady.image.tag -}}
+{{- if .Values.webhookReady.image.digest }}
+{{- $image = printf "%s@%s" $image .Values.webhookReady.image.digest -}}
+{{- end }}
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ .Release.Name }}-webhook-ready
+  namespace: {{ .Release.Namespace }}
+  annotations:
+    helm.sh/hook: post-install,post-upgrade
+    helm.sh/hook-weight: "0"
+    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  name: {{ .Release.Name }}-webhook-ready
+  annotations:
+    helm.sh/hook: post-install,post-upgrade
+    helm.sh/hook-weight: "0"
+    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
+rules:
+  - apiGroups: [""]
+    resources: ["services/proxy"]
+    resourceNames: [{{ $resourceName | quote }}]
+    verbs: ["get"]
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  name: {{ .Release.Name }}-webhook-ready
+  annotations:
+    helm.sh/hook: post-install,post-upgrade
+    helm.sh/hook-weight: "0"
+    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: {{ .Release.Name }}-webhook-ready
+subjects:
+  - kind: ServiceAccount
+    name: {{ .Release.Name }}-webhook-ready
+    namespace: {{ .Release.Namespace }}
+---
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: {{ .Release.Name }}-webhook-ready
+  namespace: {{ .Release.Namespace }}
+  annotations:
+    helm.sh/hook: post-install,post-upgrade
+    helm.sh/hook-weight: "10"
+    helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
+spec:
+  backoffLimit: {{ .Values.webhookReady.backoffLimit | default 2 }}
+  activeDeadlineSeconds: {{ $deadline }}
+  template:
+    spec:
+      restartPolicy: Never
+      serviceAccountName: {{ .Release.Name }}-webhook-ready
+      securityContext:
+        runAsNonRoot: true
+        runAsUser: 65532
+        runAsGroup: 65532
+        seccompProfile:
+          type: RuntimeDefault
+      containers:
+        - name: wait
+          image: {{ $image }}
+          imagePullPolicy: {{ if .Values.webhookReady.image.digest }}IfNotPresent{{ else }}Always{{ end }}
+          securityContext:
+            allowPrivilegeEscalation: false
+            readOnlyRootFilesystem: true
+            capabilities:
+              drop: ["ALL"]
+          command:
+            - sh
+            - -c
+            - |
+              set -e
+              ns={{ .Release.Namespace }}
+              proxy="/api/v1/namespaces/${ns}/services/{{ $resourceName }}/proxy/readyz"
+              max_attempts={{ $maxAttempts }}
+              sleep_seconds={{ $sleepSeconds }}
+              i=0
+              last_err=""
+              until last_err=$(kubectl get --raw "$proxy" 2>&1 >/dev/null); do
+                i=$((i + 1))
+                if [ $i -gt $max_attempts ]; then
+                  echo "timeout: cnpg webhook did not respond through the apiserver proxy after ${max_attempts} attempts (${sleep_seconds}s each)"
+                  echo "last error: ${last_err}"
+                  exit 1
+                fi
+                if [ $((i % 10)) -eq 1 ]; then
+                  echo "attempt $i/${max_attempts}: ${last_err}"
+                fi
+                sleep "$sleep_seconds"
+              done
+              echo "cnpg webhook is ready"
diff --git a/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
new file mode 100644
index 00000000..69f5a6b5
--- /dev/null
+++ b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
@@ -0,0 +1,308 @@
+suite: cnpg webhook post-install readiness gate
+
+templates:
+  - templates/webhook-ready-hook.yaml
+  - charts/cloudnative-pg/templates/service.yaml
+
+release:
+  name: postgres-operator
+  namespace: cozy-postgres-operator
+
+tests:
+  - it: renders four hook objects (SA + ClusterRole + ClusterRoleBinding + Job)
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - hasDocuments:
+          count: 4
+
+  - it: every rendered object carries post-install and post-upgrade hook annotations
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 0
+        equal:
+          path: metadata.annotations["helm.sh/hook"]
+          value: post-install,post-upgrade
+      - documentIndex: 1
+        equal:
+          path: metadata.annotations["helm.sh/hook"]
+          value: post-install,post-upgrade
+      - documentIndex: 2
+        equal:
+          path: metadata.annotations["helm.sh/hook"]
+          value: post-install,post-upgrade
+      - documentIndex: 3
+        equal:
+          path: metadata.annotations["helm.sh/hook"]
+          value: post-install,post-upgrade
+
+  - it: RBAC is created before the Job (hook-weight ordering)
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 0
+        equal:
+          path: kind
+          value: ServiceAccount
+      - documentIndex: 0
+        equal:
+          path: metadata.annotations["helm.sh/hook-weight"]
+          value: "0"
+      - documentIndex: 1
+        equal:
+          path: kind
+          value: ClusterRole
+      - documentIndex: 1
+        equal:
+          path: metadata.annotations["helm.sh/hook-weight"]
+          value: "0"
+      - documentIndex: 2
+        equal:
+          path: kind
+          value: ClusterRoleBinding
+      - documentIndex: 2
+        equal:
+          path: metadata.annotations["helm.sh/hook-weight"]
+          value: "0"
+      - documentIndex: 3
+        equal:
+          path: kind
+          value: Job
+      - documentIndex: 3
+        equal:
+          path: metadata.annotations["helm.sh/hook-weight"]
+          value: "10"
+
+  - it: RBAC resourceName matches the exact proxy URL segment the Job probes
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 1
+        equal:
+          path: rules[0].resources[0]
+          value: services/proxy
+      - documentIndex: 1
+        equal:
+          path: rules[0].resourceNames[0]
+          value: https:cnpg-webhook-service:webhook-server
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "/api/v1/namespaces/\\$\\{ns\\}/services/https:cnpg-webhook-service:webhook-server/proxy/readyz"
+
+  - it: hardcoded service name in the hook matches the vendored cnpg subchart Service (drift guard for make update)
+    template: charts/cloudnative-pg/templates/service.yaml
+    asserts:
+      - equal:
+          path: metadata.name
+          value: cnpg-webhook-service
+      - equal:
+          path: spec.ports[0].name
+          value: webhook-server
+
+  - it: Job calls kubectl get --raw on the proxy path
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "kubectl get --raw"
+
+  - it: Job image is digest-pinned when webhookReady.image.digest is set
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+          digest: sha256:b9ef7d8dbe65bcc81a46c09b8dc7543103055021c4f43287bf59e92a8f4fe05c
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].image
+          value: docker.io/clastix/kubectl:v1.32@sha256:b9ef7d8dbe65bcc81a46c09b8dc7543103055021c4f43287bf59e92a8f4fe05c
+
+  - it: Job image falls back to tag-only when digest is not configured
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+          digest: ""
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].image
+          value: docker.io/clastix/kubectl:v1.32
+
+  - it: backoffLimit defaults to 2 so transient pod-level failures retry instead of killing the HelmRelease
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.backoffLimit
+          value: 2
+
+  - it: backoffLimit is overridable
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        backoffLimit: 5
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.backoffLimit
+          value: 5
+
+  - it: chart render fails when webhookReady.image.repository is empty
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: ""
+          tag: v1.32
+    asserts:
+      - failedTemplate:
+          errorMessage: "webhookReady.image.repository must be set to the container image providing kubectl for the post-install readiness Job"
+
+  - it: chart render fails when webhookReady.image.tag is empty
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: ""
+    asserts:
+      - failedTemplate:
+          errorMessage: "webhookReady.image.tag must be set for the post-install readiness Job"
+
+  - it: retry loop bounds default when blanked so a wiped override still produces a working Job
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+        maxAttempts: null
+        sleepSeconds: null
+    asserts:
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "max_attempts=60"
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "sleep_seconds=2"
+
+  - it: Job pod runs non-root with seccomp RuntimeDefault for restricted-PSA clusters
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.securityContext.runAsNonRoot
+          value: true
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.securityContext.seccompProfile.type
+          value: RuntimeDefault
+
+  - it: Job container drops all capabilities and runs read-only rootfs
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].securityContext.allowPrivilegeEscalation
+          value: false
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].securityContext.readOnlyRootFilesystem
+          value: true
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].securityContext.capabilities.drop[0]
+          value: ALL
+
+  - it: imagePullPolicy is IfNotPresent when digest-pinned, Always when tag-only
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+          digest: sha256:b9ef7d8dbe65bcc81a46c09b8dc7543103055021c4f43287bf59e92a8f4fe05c
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].imagePullPolicy
+          value: IfNotPresent
+
+  - it: imagePullPolicy is Always when no digest is configured
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+          digest: ""
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].imagePullPolicy
+          value: Always
+
+  - it: retry loop captures and surfaces the last kubectl error message on timeout
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: 'last_err=\$\(kubectl get --raw'
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: 'last error: \$\{last_err\}'
+
+  - it: retry loop error message stays in sync when maxAttempts is bumped
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+        maxAttempts: 90
+        sleepSeconds: 3
+    asserts:
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "max_attempts=90"
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: "sleep_seconds=3"
+      - documentIndex: 3
+        matchRegex:
+          path: spec.template.spec.containers[0].command[2]
+          pattern: 'after \$\{max_attempts\} attempts'
+
+  - it: activeDeadlineSeconds scales with retry bounds so an override raise does not silently cut
+    template: templates/webhook-ready-hook.yaml
+    set:
+      webhookReady:
+        image:
+          repository: docker.io/clastix/kubectl
+          tag: v1.32
+        maxAttempts: 180
+        sleepSeconds: 3
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.activeDeadlineSeconds
+          value: 600
+
+  - it: activeDeadlineSeconds defaults include 60s slack over the default retry window
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.activeDeadlineSeconds
+          value: 180
diff --git a/packages/system/postgres-operator/values.yaml b/packages/system/postgres-operator/values.yaml
index cae3dc53..854a3ac6 100644
--- a/packages/system/postgres-operator/values.yaml
+++ b/packages/system/postgres-operator/values.yaml
@@ -3,3 +3,28 @@ cloudnative-pg:
     create: true
   image:
     tag: "1.27.3"
+# Image used by the post-install webhook-readiness Job (see templates/webhook-ready-hook.yaml).
+# Any image with kubectl on PATH works; the Job calls the apiserver service proxy with the
+# hook ServiceAccount's token to confirm the mcluster.cnpg.io webhook is reachable end-to-end
+# before the HelmRelease reports Ready, closing the "connection refused" bootstrap race.
+#
+# The tag is digest-pinned so an upstream retag does not change what runs on every install
+# and upgrade across the fleet. Refresh by resolving the current manifest-list digest
+# (`docker manifest inspect docker.io/clastix/kubectl:v1.32`) and updating `digest` below.
+# renovate: datasource=docker depName=docker.io/clastix/kubectl
+webhookReady:
+  image:
+    repository: docker.io/clastix/kubectl
+    tag: v1.32
+    digest: sha256:b9ef7d8dbe65bcc81a46c09b8dc7543103055021c4f43287bf59e92a8f4fe05c
+  # Retry loop bounds for the readiness probe. Defaults total ~120s wall clock.
+  # Both are `default`-coerced in the template so an override that blanks them still
+  # produces a working Job.
+  maxAttempts: 60
+  sleepSeconds: 2
+  # Pod-level retry budget for transient node/registry failures (image pull rate limit,
+  # OOM, CNI hiccup). The shell loop inside the container only covers reachability retries.
+  # Pod retries and activeDeadlineSeconds (wall-clock bound on the whole Job across all
+  # pod retries) are ANDed: a 5-minute ImagePullBackOff on attempt 0 leaves only ~60s for
+  # subsequent retries before the Job is cut.
+  backoffLimit: 2

From 648ad82dd3283076cf85f6b4b8953ceda8199476 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Fri, 24 Apr 2026 13:31:57 +0300
Subject: [PATCH 55/82] refactor(postgres-operator): scope webhook-ready RBAC
 to release namespace

Address review feedback from gemini-code-assist on
packages/system/postgres-operator/templates/webhook-ready-hook.yaml:83.

The Job targets the cnpg-webhook-service/services/proxy subresource, which
is namespaced and lives in the release namespace. A namespaced Role and
RoleBinding grant the exact permission needed without creating global
RBAC for a namespaced probe, which is the principle of least privilege.

Also update the kind assertions in tests/webhook-ready-hook_test.yaml so
the unittest suite tracks the new Role/RoleBinding objects.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../postgres-operator/templates/webhook-ready-hook.yaml   | 8 +++++---
 .../postgres-operator/tests/webhook-ready-hook_test.yaml  | 6 +++---
 2 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/packages/system/postgres-operator/templates/webhook-ready-hook.yaml b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
index 2ad4e7c9..70e51621 100644
--- a/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
+++ b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
@@ -52,9 +52,10 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
 ---
 apiVersion: rbac.authorization.k8s.io/v1
-kind: ClusterRole
+kind: Role
 metadata:
   name: {{ .Release.Name }}-webhook-ready
+  namespace: {{ .Release.Namespace }}
   annotations:
     helm.sh/hook: post-install,post-upgrade
     helm.sh/hook-weight: "0"
@@ -66,16 +67,17 @@ rules:
     verbs: ["get"]
 ---
 apiVersion: rbac.authorization.k8s.io/v1
-kind: ClusterRoleBinding
+kind: RoleBinding
 metadata:
   name: {{ .Release.Name }}-webhook-ready
+  namespace: {{ .Release.Namespace }}
   annotations:
     helm.sh/hook: post-install,post-upgrade
     helm.sh/hook-weight: "0"
     helm.sh/hook-delete-policy: before-hook-creation,hook-succeeded
 roleRef:
   apiGroup: rbac.authorization.k8s.io
-  kind: ClusterRole
+  kind: Role
   name: {{ .Release.Name }}-webhook-ready
 subjects:
   - kind: ServiceAccount
diff --git a/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
index 69f5a6b5..eb7863da 100644
--- a/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
+++ b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
@@ -9,7 +9,7 @@ release:
   namespace: cozy-postgres-operator
 
 tests:
-  - it: renders four hook objects (SA + ClusterRole + ClusterRoleBinding + Job)
+  - it: renders four hook objects (SA + Role + RoleBinding + Job)
     template: templates/webhook-ready-hook.yaml
     asserts:
       - hasDocuments:
@@ -49,7 +49,7 @@ tests:
       - documentIndex: 1
         equal:
           path: kind
-          value: ClusterRole
+          value: Role
       - documentIndex: 1
         equal:
           path: metadata.annotations["helm.sh/hook-weight"]
@@ -57,7 +57,7 @@ tests:
       - documentIndex: 2
         equal:
           path: kind
-          value: ClusterRoleBinding
+          value: RoleBinding
       - documentIndex: 2
         equal:
           path: metadata.annotations["helm.sh/hook-weight"]

From fc057c039385e7b40cca3f2b6ca77a8f80bab49a Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Fri, 24 Apr 2026 13:32:33 +0300
Subject: [PATCH 56/82] fix(postgres-operator): set resources on webhook-ready
 wait container

Address review feedback from gemini-code-assist on
packages/system/postgres-operator/templates/webhook-ready-hook.yaml:115.

The wait container had no resource requests or limits, so schedulers
treated it as BestEffort and downstream quota enforcement had no signal.
Set small requests (10m CPU, 32Mi memory) and conservative limits
(100m CPU, 64Mi memory) matching the actual footprint of the kubectl
polling loop.

Add a matching unittest assertion so the values stay in sync if anyone
touches the template.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .../templates/webhook-ready-hook.yaml         |  7 +++++++
 .../tests/webhook-ready-hook_test.yaml        | 20 +++++++++++++++++++
 2 files changed, 27 insertions(+)

diff --git a/packages/system/postgres-operator/templates/webhook-ready-hook.yaml b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
index 70e51621..51a8ba81 100644
--- a/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
+++ b/packages/system/postgres-operator/templates/webhook-ready-hook.yaml
@@ -110,6 +110,13 @@ spec:
         - name: wait
           image: {{ $image }}
           imagePullPolicy: {{ if .Values.webhookReady.image.digest }}IfNotPresent{{ else }}Always{{ end }}
+          resources:
+            requests:
+              cpu: 10m
+              memory: 32Mi
+            limits:
+              cpu: 100m
+              memory: 64Mi
           securityContext:
             allowPrivilegeEscalation: false
             readOnlyRootFilesystem: true
diff --git a/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
index eb7863da..3253c663 100644
--- a/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
+++ b/packages/system/postgres-operator/tests/webhook-ready-hook_test.yaml
@@ -205,6 +205,26 @@ tests:
           path: spec.template.spec.securityContext.seccompProfile.type
           value: RuntimeDefault
 
+  - it: wait container declares resource requests and limits for predictable scheduling
+    template: templates/webhook-ready-hook.yaml
+    asserts:
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].resources.requests.cpu
+          value: 10m
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].resources.requests.memory
+          value: 32Mi
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].resources.limits.cpu
+          value: 100m
+      - documentIndex: 3
+        equal:
+          path: spec.template.spec.containers[0].resources.limits.memory
+          value: 64Mi
+
   - it: Job container drops all capabilities and runs read-only rootfs
     template: templates/webhook-ready-hook.yaml
     asserts:

From 7da29afe6e1f33b69d6040fd1051924d7937a13b Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Fri, 24 Apr 2026 13:33:07 +0300
Subject: [PATCH 57/82] docs(postgres-operator): fix
 backoffLimit/activeDeadlineSeconds comment math

Address review feedback from coderabbitai on
packages/system/postgres-operator/values.yaml:30.

The old comment's 5-minute ImagePullBackOff scenario conflicted with
the 180s activeDeadlineSeconds that the default maxAttempts/sleepSeconds
resolve to, so the numbers could not both be taken at face value.

Rewrite the comment to state the actual deadline math and frame the two
gates as an AND with activeDeadlineSeconds being the shorter one under
defaults, so readers understand why backoffLimit has little headroom
without an accompanying maxAttempts/sleepSeconds bump.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/system/postgres-operator/values.yaml | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/packages/system/postgres-operator/values.yaml b/packages/system/postgres-operator/values.yaml
index 854a3ac6..c408220e 100644
--- a/packages/system/postgres-operator/values.yaml
+++ b/packages/system/postgres-operator/values.yaml
@@ -25,6 +25,8 @@ webhookReady:
   # Pod-level retry budget for transient node/registry failures (image pull rate limit,
   # OOM, CNI hiccup). The shell loop inside the container only covers reachability retries.
   # Pod retries and activeDeadlineSeconds (wall-clock bound on the whole Job across all
-  # pod retries) are ANDed: a 5-minute ImagePullBackOff on attempt 0 leaves only ~60s for
-  # subsequent retries before the Job is cut.
+  # pod retries) are ANDed, so activeDeadlineSeconds is the shorter of the two gates with
+  # the defaults above: the 60*2+60 = 180s deadline cuts the Job before backoffLimit=2
+  # ever matters if pod-level failures eat more than ~60s of the budget. Raise
+  # maxAttempts/sleepSeconds alongside backoffLimit when tuning for slow image pulls.
   backoffLimit: 2

From 500816b71b1ad3f916e11ff3ddb1fd6dee9fcc76 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Fri, 24 Apr 2026 16:44:46 +0300
Subject: [PATCH 58/82] refactor(ingress): move CiliumLoadBalancerIPPool to
 tenant chart
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The pool is now rendered from packages/apps/tenant/templates/cilium-lb-pool.yaml instead of packages/extra/ingress/templates/cilium-lb-pool.yaml. Cilium LB IPAM forbids overlapping CIDRs across pools regardless of serviceSelector, so both the ingress-loadBalancer path and the upcoming per-tenant Gateway in #2470 cannot each own their own pool on the same publishing.externalIPs range. The tenant chart is the natural per-tenant owner — it already creates the Namespace, the cozystack-values Secret, and the HelmReleases for both ingress and gateway.

The new pool uses a namespace-only serviceSelector (io.kubernetes.service.namespace: <ns>), which matches any LoadBalancer Service in the tenant namespace. The metadata.name changed from <trim>-ingress to <trim>-exposure to reflect that the pool is not ingress-specific.

Only the ingress-loadBalancer signal is wired in this commit (_cluster.expose-mode=loadBalancer plus .Values.ingress=true on the publishing tenant). The gateway branch is added in #2470 on top of this commit — it rebases, drops its own packages/extra/gateway/templates/cilium-lb-pool.yaml, and adds an OR branch for .Values.gateway in the tenant template.

Pool-rendering unit tests moved from packages/extra/ingress/tests/ to packages/apps/tenant/tests/. The ingress chart tests keep the Service-level asserts. packages/apps/tenant/Makefile gains a test target so hack/helm-unit-tests.sh picks up the new suite.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/tenant/Makefile                 |   3 +
 .../tenant}/templates/cilium-lb-pool.yaml     |  21 +--
 packages/apps/tenant/tests/exposure_test.yaml | 141 ++++++++++++++++++
 .../extra/ingress/tests/exposure_test.yaml    | 104 +------------
 4 files changed, 158 insertions(+), 111 deletions(-)
 rename packages/{extra/ingress => apps/tenant}/templates/cilium-lb-pool.yaml (54%)
 create mode 100644 packages/apps/tenant/tests/exposure_test.yaml

diff --git a/packages/apps/tenant/Makefile b/packages/apps/tenant/Makefile
index 3fe3810d..2f51d836 100644
--- a/packages/apps/tenant/Makefile
+++ b/packages/apps/tenant/Makefile
@@ -3,3 +3,6 @@ include ../../../hack/package.mk
 generate:
 	cozyvalues-gen -m 'tenant' -v values.yaml -s values.schema.json -r README.md -g ../../../api/apps/v1alpha1/tenant/types.go
 	../../../hack/update-crd.sh
+
+test:
+	helm unittest .
diff --git a/packages/extra/ingress/templates/cilium-lb-pool.yaml b/packages/apps/tenant/templates/cilium-lb-pool.yaml
similarity index 54%
rename from packages/extra/ingress/templates/cilium-lb-pool.yaml
rename to packages/apps/tenant/templates/cilium-lb-pool.yaml
index 048eb58b..63dbcaca 100644
--- a/packages/extra/ingress/templates/cilium-lb-pool.yaml
+++ b/packages/apps/tenant/templates/cilium-lb-pool.yaml
@@ -1,25 +1,28 @@
-{{- $exposeIngress := (index .Values._cluster "expose-ingress") | default "tenant-root" }}
 {{- $exposeMode := (index .Values._cluster "expose-mode") | default "externalIPs" }}
-{{- $exposeExternalIPs := (index .Values._cluster "expose-external-ips") | default "" | nospace }}
-{{- $exposeIPsList := list }}
-{{- range splitList "," $exposeExternalIPs }}
+{{- $exposeIngress := (index .Values._cluster "expose-ingress") | default "tenant-root" }}
+{{- $exposeIPs := (index .Values._cluster "expose-external-ips") | default "" | nospace }}
+{{- $ipsList := list }}
+{{- range splitList "," $exposeIPs }}
   {{- $ip := . | trim }}
   {{- if $ip }}
-    {{- $exposeIPsList = append $exposeIPsList $ip }}
+    {{- $ipsList = append $ipsList $ip }}
   {{- end }}
 {{- end }}
-{{- if and (eq $exposeMode "loadBalancer") (eq $exposeIngress .Release.Namespace) $exposeIPsList }}
+{{- $isPublishingIngressLB := and
+      (eq $exposeMode "loadBalancer")
+      (eq $exposeIngress .Release.Namespace)
+      .Values.ingress }}
+{{- if and $isPublishingIngressLB $ipsList }}
 apiVersion: cilium.io/v2
 kind: CiliumLoadBalancerIPPool
 metadata:
-  name: {{ trimPrefix "tenant-" .Release.Namespace }}-ingress
+  name: {{ trimPrefix "tenant-" .Release.Namespace }}-exposure
 spec:
   blocks:
-  {{- range $exposeIPsList }}
+  {{- range $ipsList }}
   - cidr: {{ . }}{{ if not (contains "/" .) }}/{{ if contains ":" . }}128{{ else }}32{{ end }}{{ end }}
   {{- end }}
   serviceSelector:
     matchLabels:
       "io.kubernetes.service.namespace": {{ .Release.Namespace | quote }}
-      "app.kubernetes.io/name": ingress-nginx
 {{- end }}
diff --git a/packages/apps/tenant/tests/exposure_test.yaml b/packages/apps/tenant/tests/exposure_test.yaml
new file mode 100644
index 00000000..e86fab5b
--- /dev/null
+++ b/packages/apps/tenant/tests/exposure_test.yaml
@@ -0,0 +1,141 @@
+suite: tenant CiliumLoadBalancerIPPool rendering for publishing.exposure=loadBalancer
+templates:
+  - templates/cilium-lb-pool.yaml
+
+release:
+  name: tenant-root
+  namespace: tenant-root
+
+tests:
+  - it: default exposure (externalIPs) renders no pool
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,192.0.2.11"
+    asserts:
+      - hasDocuments:
+          count: 0
+
+  - it: loadBalancer mode in publishing tenant with ingress=true renders v2 pool with namespace-only selector
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,192.0.2.11"
+        expose-mode: loadBalancer
+    asserts:
+      - hasDocuments:
+          count: 1
+      - equal:
+          path: apiVersion
+          value: cilium.io/v2
+      - equal:
+          path: kind
+          value: CiliumLoadBalancerIPPool
+      - equal:
+          path: metadata.name
+          value: root-exposure
+      - equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 192.0.2.11/32
+      - equal:
+          path: spec.serviceSelector.matchLabels["io.kubernetes.service.namespace"]
+          value: tenant-root
+      - notExists:
+          path: spec.serviceSelector.matchLabels["app.kubernetes.io/name"]
+
+  - it: loadBalancer mode with IPv6 emits /128 CIDR
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "2001:db8::1"
+        expose-mode: loadBalancer
+    asserts:
+      - equal:
+          path: spec.blocks
+          value:
+            - cidr: 2001:db8::1/128
+
+  - it: loadBalancer mode with mixed IPv4 and IPv6 emits correct CIDR per family
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,2001:db8::1"
+        expose-mode: loadBalancer
+    asserts:
+      - equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 2001:db8::1/128
+
+  - it: loadBalancer mode accepts pre-CIDR input without double-suffixing
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10/32,2001:db8::1/128"
+        expose-mode: loadBalancer
+    asserts:
+      - equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 2001:db8::1/128
+
+  - it: loadBalancer mode filters out empty entries from externalIPs (trailing, leading, repeated commas)
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10,,192.0.2.11,"
+        expose-mode: loadBalancer
+    asserts:
+      - hasDocuments:
+          count: 1
+      - equal:
+          path: spec.blocks
+          value:
+            - cidr: 192.0.2.10/32
+            - cidr: 192.0.2.11/32
+
+  - it: loadBalancer mode with ingress=false in publishing tenant renders no pool
+    set:
+      ingress: false
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+        expose-mode: loadBalancer
+    asserts:
+      - hasDocuments:
+          count: 0
+
+  - it: loadBalancer mode in a non-publishing tenant renders no pool
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: "192.0.2.10"
+        expose-mode: loadBalancer
+    release:
+      name: tenant-u1
+      namespace: tenant-u1
+    asserts:
+      - hasDocuments:
+          count: 0
+
+  - it: loadBalancer mode in publishing tenant with empty externalIPs renders no pool
+    set:
+      ingress: true
+      _cluster:
+        expose-ingress: tenant-root
+        expose-external-ips: ""
+        expose-mode: loadBalancer
+    asserts:
+      - hasDocuments:
+          count: 0
diff --git a/packages/extra/ingress/tests/exposure_test.yaml b/packages/extra/ingress/tests/exposure_test.yaml
index 323c8e68..1c76b3da 100644
--- a/packages/extra/ingress/tests/exposure_test.yaml
+++ b/packages/extra/ingress/tests/exposure_test.yaml
@@ -1,14 +1,13 @@
 suite: ingress exposure modes
 templates:
   - templates/nginx-ingress.yaml
-  - templates/cilium-lb-pool.yaml
 
 release:
   name: ingress
   namespace: tenant-root
 
 tests:
-  - it: default exposure (externalIPs) renders ClusterIP Service with spec.externalIPs and no CiliumLoadBalancerIPPool
+  - it: default exposure (externalIPs) renders ClusterIP Service with spec.externalIPs
     set:
       _cluster:
         expose-ingress: tenant-root
@@ -31,9 +30,6 @@ tests:
       - template: templates/nginx-ingress.yaml
         notExists:
           path: spec.values.ingress-nginx.controller.service.labels
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 0
 
   - it: legacy config without expose-mode falls back to externalIPs behavior
     set:
@@ -45,9 +41,6 @@ tests:
         equal:
           path: spec.values.ingress-nginx.controller.service.type
           value: ClusterIP
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 0
 
   - it: externalIPs mode in a namespace other than publishing.ingressName renders LoadBalancer fallback without externalIPs
     set:
@@ -68,11 +61,8 @@ tests:
       - template: templates/nginx-ingress.yaml
         notExists:
           path: spec.values.ingress-nginx.controller.service.externalIPs
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 0
 
-  - it: loadBalancer mode renders LoadBalancer Service with lb-pool label and a v2 CiliumLoadBalancerIPPool
+  - it: loadBalancer mode renders LoadBalancer Service without externalIPs on the Service
     set:
       _cluster:
         expose-ingress: tenant-root
@@ -93,62 +83,6 @@ tests:
       - template: templates/nginx-ingress.yaml
         notExists:
           path: spec.values.ingress-nginx.controller.service.externalIPs
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 1
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: apiVersion
-          value: cilium.io/v2
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: kind
-          value: CiliumLoadBalancerIPPool
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: metadata.name
-          value: root-ingress
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.blocks
-          value:
-            - cidr: 192.0.2.10/32
-            - cidr: 192.0.2.11/32
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.serviceSelector.matchLabels["io.kubernetes.service.namespace"]
-          value: tenant-root
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.serviceSelector.matchLabels["app.kubernetes.io/name"]
-          value: ingress-nginx
-
-  - it: loadBalancer mode with IPv6 address emits /128 CIDR
-    set:
-      _cluster:
-        expose-ingress: tenant-root
-        expose-external-ips: "2001:db8::1"
-        expose-mode: loadBalancer
-    asserts:
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.blocks
-          value:
-            - cidr: 2001:db8::1/128
-
-  - it: loadBalancer mode with mixed IPv4 and IPv6 emits correct CIDR per family
-    set:
-      _cluster:
-        expose-ingress: tenant-root
-        expose-external-ips: "192.0.2.10,2001:db8::1"
-        expose-mode: loadBalancer
-    asserts:
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.blocks
-          value:
-            - cidr: 192.0.2.10/32
-            - cidr: 2001:db8::1/128
 
   - it: loadBalancer mode without externalIPs fails chart render with explicit message
     set:
@@ -206,26 +140,6 @@ tests:
       - template: templates/nginx-ingress.yaml
         notExists:
           path: spec.values.ingress-nginx.controller.service.externalIPs
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 0
-
-  - it: loadBalancer mode filters out empty entries from externalIPs (trailing comma, leading comma)
-    set:
-      _cluster:
-        expose-ingress: tenant-root
-        expose-external-ips: "192.0.2.10,,192.0.2.11,"
-        expose-mode: loadBalancer
-    asserts:
-      - template: templates/cilium-lb-pool.yaml
-        hasDocuments:
-          count: 1
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.blocks
-          value:
-            - cidr: 192.0.2.10/32
-            - cidr: 192.0.2.11/32
 
   - it: loadBalancer mode with only-empty externalIPs fails chart render (comma-only input)
     set:
@@ -249,17 +163,3 @@ tests:
           path: spec.values.ingress-nginx.controller.service.externalIPs
           value:
             - 192.0.2.10
-
-  - it: loadBalancer mode accepts pre-CIDR input without double-suffixing
-    set:
-      _cluster:
-        expose-ingress: tenant-root
-        expose-external-ips: "192.0.2.10/32,2001:db8::1/128"
-        expose-mode: loadBalancer
-    asserts:
-      - template: templates/cilium-lb-pool.yaml
-        equal:
-          path: spec.blocks
-          value:
-            - cidr: 192.0.2.10/32
-            - cidr: 2001:db8::1/128

From c8ed1c652cbd3d442036251358c68e7b25ec64a1 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 02:35:52 +0300
Subject: [PATCH 61/82] chore(ci): adopt CNCF/k8s label conventions
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Add .github/labels.yml as the canonical label set, synced into the
repository by .github/workflows/labels.yaml using EndBug/label-sync.

Conventions follow the Kubernetes scheme:
https://github.com/kubernetes/test-infra/blob/master/label_sync/labels.md

Six namespaced groups: kind/, priority/, triage/, lifecycle/, area/,
do-not-merge/. Cozystack-specific labels preserved (epic, community,
security/*, size:*).

Migration via aliases keeps references on existing issues and PRs:
- bug           -> kind/bug
- enhancement   -> kind/feature
- documentation -> kind/documentation
- question      -> kind/support
- frozen        -> lifecycle/frozen
- stale         -> lifecycle/stale
- do-not-merge  -> do-not-merge/work-in-progress

delete-other-labels is false on the initial rollout; redundant labels
("do not merge", duplicate, invalid, wontfix) stay until a follow-up
PR removes them after stabilisation.

The labels workflow has a validate job (python3 schema check) that
runs on PR. Sync runs only on push to main, weekly cron, and manual
dispatch. Schema invariants:
- description <= 100 chars (GitHub REST API limit)
- color is 6-char hex without leading #
- unique top-level names
- aliases do not collide with top-level names

PR auto-labeling (.github/workflows/pr-labeler.yaml):
- Parses PR title as Conventional Commits header (type, scope, !).
- type -> kind/* (feat -> kind/feature, fix -> kind/bug, docs ->
  kind/documentation, chore/refactor -> kind/cleanup; style, perf,
  test, build, ci, revert -> no kind label).
- scope -> area/* via embedded mapping; composite scopes split on
  comma. Bracket-style fallback ([scope] description) maps area/*
  but cannot infer kind/*.
- '[Backport release-1.x]' prefix is stripped; area/release and
  backport labels are added.
- '!' after type or 'BREAKING CHANGE:' footer in body adds
  kind/breaking-change.
- Unmapped scope or non-conventional title adds area/uncategorized
  to flag for human review.
- Additive only — never removes existing labels.

Hardcoded label references updated:
- .github/ISSUE_TEMPLATE/bug_report.md (bug -> kind/bug)
- .github/workflows/tags.yaml (documentation -> kind/documentation)

AGENTS.md gains an Activation entry pointing agents to labels.yml
as the source of truth and to contributing.md for the title
auto-labeling table.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/ISSUE_TEMPLATE/bug_report.md |   2 +-
 .github/labels.yml                   | 371 +++++++++++++++++++++++++++
 .github/workflows/labels.yaml        |  84 ++++++
 .github/workflows/pr-labeler.yaml    | 199 ++++++++++++++
 .github/workflows/tags.yaml          |   6 +-
 AGENTS.md                            |   6 +
 docs/agents/contributing.md          |  43 ++++
 7 files changed, 707 insertions(+), 4 deletions(-)
 create mode 100644 .github/labels.yml
 create mode 100644 .github/workflows/labels.yaml
 create mode 100644 .github/workflows/pr-labeler.yaml

diff --git a/.github/ISSUE_TEMPLATE/bug_report.md b/.github/ISSUE_TEMPLATE/bug_report.md
index 1a119259..fefdab82 100644
--- a/.github/ISSUE_TEMPLATE/bug_report.md
+++ b/.github/ISSUE_TEMPLATE/bug_report.md
@@ -1,7 +1,7 @@
 ---
 name: Bug report
 about: Create a report to help us improve
-labels: 'bug'
+labels: 'kind/bug'
 assignees: ''
 
 ---
diff --git a/.github/labels.yml b/.github/labels.yml
new file mode 100644
index 00000000..1369aa48
--- /dev/null
+++ b/.github/labels.yml
@@ -0,0 +1,371 @@
+# Cozystack repository labels
+#
+# Label conventions follow the Kubernetes scheme:
+# https://github.com/kubernetes/test-infra/blob/master/label_sync/labels.md
+#
+# Synced into the repository by .github/workflows/labels.yaml
+# (EndBug/label-sync@v2). Edit this file via pull request — UI changes
+# will be overwritten on the next sync.
+#
+# Constraints (enforced by the validate job in labels.yaml):
+#   - description ≤ 100 characters (GitHub REST API limit)
+#   - color is a 6-character hex string (no leading #)
+#   - label names are unique
+#   - aliases do not collide with top-level names
+#
+# Categories:
+#   kind/         issue or PR type
+#   priority/     urgency
+#   triage/       review state
+#   lifecycle/    issue or PR lifecycle
+#   area/         subsystem; extensible — add when 3+ open issues exist
+#   do-not-merge/ PR merge blockers
+#   security/     security-finding severity and status (Cozystack-specific)
+#   size:         PR size (auto-applied)
+#
+# `aliases:` lets EndBug/label-sync rename existing labels without losing
+# references on already-tagged issues and PRs.
+#
+# GitHub-default labels not migrated here (`wontfix`, `invalid`) currently
+# carry zero issues/PRs in this repo and will be removed in a follow-up
+# cleanup PR rather than aliased to a different namespace.
+
+# ──────────────────────────────────────────────
+# kind/ — issue or PR type
+# ──────────────────────────────────────────────
+
+- name: kind/bug
+  color: 'd73a4a'
+  description: Categorizes issue or PR as related to a bug
+  aliases: ['bug']
+
+- name: kind/feature
+  color: 'a2eeef'
+  description: Categorizes issue or PR as related to a new feature
+  aliases: ['enhancement']
+
+- name: kind/documentation
+  color: '0075ca'
+  description: Categorizes issue or PR as related to documentation
+  aliases: ['documentation']
+
+- name: kind/support
+  color: 'd876e3'
+  description: Categorizes issue as a support question
+  aliases: ['question']
+
+- name: kind/cleanup
+  color: 'c7def8'
+  description: Categorizes issue or PR as related to cleanup of code, process, or technical debt
+
+- name: kind/regression
+  color: 'e11d21'
+  description: Categorizes issue or PR as related to a regression from a prior release
+
+- name: kind/flake
+  color: 'f7c6c7'
+  description: Categorizes issue or PR as related to a flaky test
+
+- name: kind/failing-test
+  color: 'e11d21'
+  description: Categorizes issue or PR as related to a consistently or frequently failing test
+
+- name: kind/api-change
+  color: 'c7def8'
+  description: Categorizes issue or PR as related to adding, removing, or otherwise changing an API
+
+- name: kind/breaking-change
+  color: 'e11d21'
+  description: Indicates the change introduces a breaking API or behaviour change
+
+# ──────────────────────────────────────────────
+# priority/ — urgency
+# ──────────────────────────────────────────────
+
+- name: priority/critical-urgent
+  color: 'e11d21'
+  description: Highest priority. Must be actively worked on as someone's top priority right now
+
+- name: priority/important-soon
+  color: 'eb6420'
+  description: Must be staffed and worked on either currently, or very soon, ideally in time for the next release
+
+- name: priority/important-longterm
+  color: 'fbca04'
+  description: Important over the long term, but may not be staffed and/or may need multiple releases to complete
+
+- name: priority/backlog
+  color: 'fef2c0'
+  description: General backlog priority. Lower than priority/important-longterm
+
+# ──────────────────────────────────────────────
+# triage/ — review state
+# ──────────────────────────────────────────────
+
+- name: triage/needs-triage
+  color: 'ededed'
+  description: Indicates an issue needs triage by a maintainer
+
+- name: triage/accepted
+  color: '0e8a16'
+  description: Indicates an issue is ready to be actively worked on
+
+- name: triage/needs-information
+  color: 'fbca04'
+  description: Indicates an issue needs more information in order to work on it
+
+- name: triage/not-reproducible
+  color: 'fbca04'
+  description: Indicates an issue can not be reproduced as described
+
+- name: triage/duplicate
+  color: 'cfd3d7'
+  description: Indicates an issue is a duplicate of another issue
+  aliases: ['duplicate']
+
+- name: triage/unresolved
+  color: 'cfd3d7'
+  description: Indicates an issue that can not or will not be resolved
+
+# ──────────────────────────────────────────────
+# lifecycle/ — issue or PR lifecycle
+# ──────────────────────────────────────────────
+
+- name: lifecycle/active
+  color: '1d76db'
+  description: Indicates that an issue or PR is actively being worked on by a contributor
+
+- name: lifecycle/frozen
+  color: 'db5dd6'
+  description: Indicates that an issue or PR should not be auto-closed due to staleness
+  aliases: ['frozen']
+
+- name: lifecycle/stale
+  color: 'dadada'
+  description: Denotes an issue or PR has remained open with no activity and has become stale
+  aliases: ['stale']
+
+- name: lifecycle/rotten
+  color: '795548'
+  description: Denotes an issue or PR that has aged beyond stale and will be auto-closed
+
+# ──────────────────────────────────────────────
+# area/ — subsystem (extensible)
+# Add a new area/* when there are 3+ open issues on the topic.
+# ──────────────────────────────────────────────
+
+- name: area/api
+  color: 'bfd4f2'
+  description: Issues or PRs related to the cozystack-api aggregated API server
+
+- name: area/ai
+  color: 'bfd4f2'
+  description: Issues or PRs related to AI agent guides, AGENTS.md, docs/agents/
+
+- name: area/build
+  color: 'bfd4f2'
+  description: Issues or PRs related to image build infrastructure, multi-arch support
+
+- name: area/ci
+  color: 'bfd4f2'
+  description: Issues or PRs related to CI workflows, GitHub Actions, automation
+
+- name: area/dashboard
+  color: 'bfd4f2'
+  description: Issues or PRs related to the dashboard / UI
+
+- name: area/extra
+  color: 'bfd4f2'
+  description: Issues or PRs related to tenant-specific modules (packages/extra/)
+
+- name: area/database
+  color: 'bfd4f2'
+  description: Issues or PRs related to managed databases (postgres, mariadb, redis, etcd, kafka, clickhouse)
+
+- name: area/kubernetes
+  color: 'bfd4f2'
+  description: Issues or PRs related to the tenant Kubernetes app
+
+- name: area/monitoring
+  color: 'bfd4f2'
+  description: Issues or PRs related to the monitoring stack (vlogs, vmstack, grafana, workloadmonitor)
+
+- name: area/networking
+  color: 'bfd4f2'
+  description: Issues or PRs related to networking (ingress, gateway, vpn, metallb, cilium, kube-ovn)
+
+- name: area/platform
+  color: 'bfd4f2'
+  description: Issues or PRs related to platform infrastructure (bundle, flux, talos, installer)
+
+- name: area/release
+  color: 'bfd4f2'
+  description: Issues or PRs related to release tooling (changelog, backport, release pipeline)
+
+- name: area/storage
+  color: 'bfd4f2'
+  description: Issues or PRs related to storage (linstor, seaweedfs, bucket, velero, harbor)
+
+- name: area/testing
+  color: 'bfd4f2'
+  description: Issues or PRs related to testing (e2e, bats, unit tests)
+
+- name: area/virtualization
+  color: 'bfd4f2'
+  description: Issues or PRs related to virtualization (kubevirt, cdi, vmi, vm-import)
+
+- name: area/uncategorized
+  color: 'fbca04'
+  description: PR auto-labeler could not map title scope to a known area/*; please review
+
+# ──────────────────────────────────────────────
+# do-not-merge/ — PR merge blockers (Prow convention)
+# ──────────────────────────────────────────────
+
+- name: do-not-merge/work-in-progress
+  color: 'e11d21'
+  description: Indicates that a PR should not merge because it is a work in progress
+  # Both legacy spellings collapse here. EndBug processes aliases sequentially;
+  # the second rename hits a name collision and logs a warning — the legacy
+  # label survives and gets cleaned up in the follow-up dedup PR.
+  aliases: ['do-not-merge', 'do not merge']
+
+- name: do-not-merge/hold
+  color: 'e11d21'
+  description: Indicates that a PR should not merge because someone has issued /hold
+
+# ──────────────────────────────────────────────
+# Cozystack-specific (preserved)
+# ──────────────────────────────────────────────
+
+- name: epic
+  color: 'A335EE'
+  description: A large development increment that brings definite value to Cozystack users
+
+- name: community
+  color: '97458A'
+  description: Community contributions are welcome in this issue
+
+- name: help wanted
+  color: '008672'
+  description: Extra attention is needed
+
+- name: good first issue
+  color: '7057ff'
+  description: Good for newcomers
+
+- name: quality-of-life
+  color: 'aaaaaa'
+  description: QoL improvements
+
+- name: upstream-issue
+  color: 'aaaaaa'
+  description: Requires resolving an issue in an upstream project
+
+- name: backport
+  color: 'FBCA04'
+  description: Should change be backported on previous release
+
+- name: backport-previous
+  color: 'fbd876'
+  description: Backport target — previous release line
+
+- name: release
+  color: 'aaaaaa'
+  description: Releasing a new Cozystack version
+
+- name: automated
+  color: 'ededed'
+  description: Created by automation
+
+- name: debug
+  color: '704479'
+  description: Debugging in progress
+
+- name: sponsored
+  color: '00FF00'
+  description: Sponsored work
+
+- name: lgtm
+  color: '238636'
+  description: This PR has been approved by a maintainer
+
+- name: ok-to-test
+  color: '00FF00'
+  description: Indicates a non-member PR is safe to run CI on
+
+# ──────────────────────────────────────────────
+# size: — PR size (auto-applied by sizing bot)
+# ──────────────────────────────────────────────
+
+- name: 'size:XS'
+  color: '00ff00'
+  description: This PR changes 0-9 lines, ignoring generated files
+
+- name: 'size:S'
+  color: '77b800'
+  description: This PR changes 10-29 lines, ignoring generated files
+
+- name: 'size:M'
+  color: 'ebb800'
+  description: This PR changes 30-99 lines, ignoring generated files
+
+- name: 'size:L'
+  color: 'eb9500'
+  description: This PR changes 100-499 lines, ignoring generated files
+
+- name: 'size:XL'
+  color: 'ff823f'
+  description: This PR changes 500-999 lines, ignoring generated files
+
+- name: 'size:XXL'
+  color: 'ffb8b8'
+  description: This PR changes 1000+ lines, ignoring generated files
+
+# ──────────────────────────────────────────────
+# security/ — security-finding severity and status
+# ──────────────────────────────────────────────
+
+- name: security
+  color: 'aaaaaa'
+  description: Security-related issues and features
+
+- name: security/critical
+  color: 'd73a4a'
+  description: Critical security vulnerability
+
+- name: security/high
+  color: 'e99695'
+  description: High severity security finding
+
+- name: security/medium
+  color: 'f9c513'
+  description: Medium severity security finding
+
+- name: security/low
+  color: '0e8a16'
+  description: Low severity security finding
+
+- name: security/triage-needed
+  color: 'fbca04'
+  description: Needs security triage
+
+- name: security/confirmed
+  color: '1d76db'
+  description: Confirmed vulnerability
+
+- name: security/false-positive
+  color: 'c5def5'
+  description: Triaged as false positive
+
+- name: security/accepted-risk
+  color: 'bfd4f2'
+  description: Risk accepted with justification
+
+- name: security/in-progress
+  color: '0075ca'
+  description: Fix in progress
+
+- name: security/fixed
+  color: '0e8a16'
+  description: Fix released
diff --git a/.github/workflows/labels.yaml b/.github/workflows/labels.yaml
new file mode 100644
index 00000000..69c18329
--- /dev/null
+++ b/.github/workflows/labels.yaml
@@ -0,0 +1,84 @@
+name: Labels
+
+on:
+  pull_request:
+    paths:
+      - .github/labels.yml
+      - .github/workflows/labels.yaml
+  push:
+    branches: [main]
+    paths:
+      - .github/labels.yml
+      - .github/workflows/labels.yaml
+  workflow_dispatch:
+  schedule:
+    - cron: '17 4 * * 1'  # Mondays at 04:17 UTC
+
+permissions:
+  contents: read
+
+concurrency:
+  group: labels-sync
+  cancel-in-progress: false
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Validate labels.yml schema
+        run: |
+          python3 - <<'PY'
+          import re, sys, yaml
+
+          path = '.github/labels.yml'
+          data = yaml.safe_load(open(path))
+          errors = []
+
+          # 1. description ≤ 100 chars (GitHub REST API limit)
+          for label in data:
+              desc = label.get('description', '') or ''
+              if len(desc) > 100:
+                  errors.append(f"{label['name']}: description {len(desc)} chars (max 100)")
+
+          # 2. color is 6-char hex without leading #
+          for label in data:
+              color = label.get('color', '') or ''
+              if not re.match(r'^[0-9A-Fa-f]{6}$', color):
+                  errors.append(f"{label['name']}: bad color {color!r} (must be 6-char hex without #)")
+
+          # 3. unique top-level names
+          names = [label['name'] for label in data]
+          dups = sorted({n for n in names if names.count(n) > 1})
+          for n in dups:
+              errors.append(f"duplicate name: {n}")
+
+          # 4. aliases do not collide with any top-level name
+          name_set = set(names)
+          for label in data:
+              for alias in (label.get('aliases') or []):
+                  if alias in name_set:
+                      errors.append(f"alias {alias!r} (under {label['name']!r}) collides with a top-level name")
+
+          if errors:
+              for err in errors:
+                  print(f"::error::{err}")
+              sys.exit(1)
+
+          print(f"labels.yml schema OK ({len(data)} labels)")
+          PY
+
+  sync:
+    needs: validate
+    if: github.event_name != 'pull_request'
+    runs-on: ubuntu-latest
+    permissions:
+      issues: write
+      pull-requests: write
+    steps:
+      - uses: actions/checkout@v4
+      - uses: EndBug/label-sync@v2
+        with:
+          config-file: .github/labels.yml
+          delete-other-labels: false
diff --git a/.github/workflows/pr-labeler.yaml b/.github/workflows/pr-labeler.yaml
new file mode 100644
index 00000000..d882bc50
--- /dev/null
+++ b/.github/workflows/pr-labeler.yaml
@@ -0,0 +1,199 @@
+name: PR Auto-Label
+
+on:
+  pull_request_target:
+    types: [opened, edited, reopened, synchronize]
+
+permissions:
+  contents: read
+  pull-requests: write
+
+jobs:
+  label:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Apply labels from PR title
+        uses: actions/github-script@v7
+        with:
+          script: |
+            // Conventional Commits types accepted by Cozystack (per docs/agents/contributing.md):
+            //   feat, fix, docs, style, refactor, perf, test, build, ci, chore, revert
+            // Mapping below maps a subset to kind/* — types not listed do not produce a kind/*.
+            const typeToKind = {
+              feat:     'kind/feature',
+              fix:      'kind/bug',
+              docs:     'kind/documentation',
+              chore:    'kind/cleanup',
+              refactor: 'kind/cleanup',
+              // style, perf, test, build, ci, revert — no kind mapping
+            };
+
+            // scope -> area/* mapping. Keys are the scopes observed in cozystack issues
+            // and PRs. Add new entries when a scope recurs (3+ times).
+            const scopeToArea = {
+              // area/api
+              'api':                 'area/api',
+              'cozystack-api':       'area/api',
+
+              // area/ai
+              'agents':              'area/ai',
+              'ai':                  'area/ai',
+
+              // area/build
+              'build':               'area/build',
+
+              // area/ci
+              'ci':                  'area/ci',
+
+              // area/dashboard
+              'dashboard':           'area/dashboard',
+
+              // area/database
+              'postgres':            'area/database',
+              'postgres-operator':   'area/database',
+              'mariadb':             'area/database',
+              'mariadb-operator':    'area/database',
+              'redis':               'area/database',
+              'etcd':                'area/database',
+              'kafka':               'area/database',
+              'clickhouse':          'area/database',
+
+              // area/extra
+              'extra':               'area/extra',
+
+              // area/kubernetes
+              'kubernetes':          'area/kubernetes',
+              'apps/kubernetes':     'area/kubernetes',
+
+              // area/monitoring
+              'monitoring':          'area/monitoring',
+              'vlogs':               'area/monitoring',
+              'vmstack':             'area/monitoring',
+              'grafana':             'area/monitoring',
+              'workloadmonitor':     'area/monitoring',
+
+              // area/networking
+              'ingress':             'area/networking',
+              'ingress-nginx':       'area/networking',
+              'gateway':             'area/networking',
+              'vpn':                 'area/networking',
+              'metallb':             'area/networking',
+              'cilium':              'area/networking',
+              'kube-ovn':            'area/networking',
+              'tcp-balancer':        'area/networking',
+              'securitygroups':      'area/networking',
+              'cozy-proxy':          'area/networking',
+
+              // area/platform
+              'platform':            'area/platform',
+              'bundle':              'area/platform',
+              'flux':                'area/platform',
+              'fluxcd':              'area/platform',
+              'cluster-api':         'area/platform',
+              'talos':               'area/platform',
+              'installer':           'area/platform',
+              'cozyctl':             'area/platform',
+              'cozystack-engine':    'area/platform',
+              'cozy-lib':            'area/platform',
+
+              // area/release
+              'backport':            'area/release',
+              'release':             'area/release',
+
+              // area/storage
+              'seaweedfs':           'area/storage',
+              'seaweedfs-cosi-driver': 'area/storage',
+              'bucket':              'area/storage',
+              'linstor':             'area/storage',
+              'velero':              'area/storage',
+              'harbor':              'area/storage',
+              'backups':             'area/storage',
+
+              // area/testing
+              'tests':               'area/testing',
+              'e2e':                 'area/testing',
+
+              // area/virtualization
+              'kubevirt':            'area/virtualization',
+              'cdi':                 'area/virtualization',
+              'vmi':                 'area/virtualization',
+              'vm-import':           'area/virtualization',
+              'virtual-machine':     'area/virtualization',
+              'hami':                'area/virtualization',
+              'gpu-operator':        'area/virtualization',
+            };
+
+            const pr = context.payload.pull_request;
+            const title = pr.title || '';
+            const body  = pr.body  || '';
+            const existing = new Set(pr.labels.map(l => l.name));
+            const toAdd = new Set();
+
+            // 1. Strip "[Backport release-1.x]" prefix if present.
+            const backportMatch = title.match(/^\[Backport ([^\]]+)\]\s+(.+)$/);
+            const cleanTitle = backportMatch ? backportMatch[2] : title;
+            if (backportMatch) {
+              toAdd.add('area/release');
+              toAdd.add('backport');
+            }
+
+            // 2. Try Conventional Commits form: type(scope)?(!)?: description
+            const conv = cleanTitle.match(/^([a-z]+)(?:\(([^)]+)\))?(!)?:\s*.+$/);
+            // 3. Fall back to bracket form: [scope] description
+            const bracket = !conv && cleanTitle.match(/^\[([^\]]+)\]\s+.+$/);
+
+            let type = null, scopeStr = null, breaking = false;
+            if (conv) {
+              type = conv[1];
+              scopeStr = conv[2] || null;
+              breaking = !!conv[3];
+            } else if (bracket) {
+              scopeStr = bracket[1];
+            }
+
+            // 4. Detect BREAKING CHANGE: footer in body.
+            if (/^BREAKING CHANGE:/m.test(body)) {
+              breaking = true;
+            }
+
+            // 5. Apply kind/* from type.
+            if (type && typeToKind[type]) {
+              toAdd.add(typeToKind[type]);
+            }
+
+            // 6. Apply area/* from scope. Composite scopes split on comma.
+            const scopes = (scopeStr || '')
+              .split(/,\s*/)
+              .map(s => s.trim())
+              .filter(Boolean);
+            for (const s of scopes) {
+              if (scopeToArea[s]) {
+                toAdd.add(scopeToArea[s]);
+              }
+            }
+
+            // 7. kind/breaking-change.
+            if (breaking) {
+              toAdd.add('kind/breaking-change');
+            }
+
+            // 8. Fallback: no area/* applied -> area/uncategorized.
+            const hasArea = [...toAdd].some(l => l.startsWith('area/'));
+            if (!hasArea) {
+              toAdd.add('area/uncategorized');
+            }
+
+            // 9. Additive only — never remove existing labels.
+            const newLabels = [...toAdd].filter(l => !existing.has(l));
+            if (newLabels.length === 0) {
+              core.info('No new labels to apply');
+              return;
+            }
+
+            await github.rest.issues.addLabels({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: pr.number,
+              labels: newLabels,
+            });
+            core.info(`Applied labels: ${newLabels.join(', ')}`);
diff --git a/.github/workflows/tags.yaml b/.github/workflows/tags.yaml
index 90f04829..1b68ee32 100644
--- a/.github/workflows/tags.yaml
+++ b/.github/workflows/tags.yaml
@@ -223,7 +223,7 @@ jobs:
                 repo:  context.repo.repo,
                 head,
                 base,
-                title: `Release v${version}`,
+                title: `chore(release): cut v${version}`,
                 body:  `This PR prepares the release \`v${version}\`.`,
                 draft: false
               });
@@ -411,7 +411,7 @@ jobs:
                 repo:  context.repo.repo,
                 head:  changelogBranch,
                 base:  baseBranch,
-                title: `docs: add changelog for v${version}`,
+                title: `docs(release): add changelog for v${version}`,
                 body:  `This PR adds the changelog for release \`v${version}\`.\n\n✅ Changelog has been automatically generated in \`docs/changelogs/v${version}.md\`.`,
                 draft: false
               });
@@ -421,7 +421,7 @@ jobs:
                 owner: context.repo.owner,
                 repo:  context.repo.repo,
                 issue_number: pr.data.number,
-                labels: ['documentation', 'automated']
+                labels: ['kind/documentation', 'automated']
               });
               
               console.log(`Created PR #${pr.data.number} for changelog`);
diff --git a/AGENTS.md b/AGENTS.md
index fe9aa8e1..5298f6fd 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -27,6 +27,12 @@ working with the **Cozystack** project.
   - Read: [`contributing.md`](./docs/agents/contributing.md)
   - Action: Read the file to understand git workflow, commit format, PR process
 
+- **Issue and PR labeling, triage** (e.g., "label this issue", "what label should I use", "triage this", "categorize")
+  - Read: [`.github/labels.yml`](./.github/labels.yml)
+  - Action: Use labels defined there. Conventions follow the Kubernetes scheme — `kind/*` (type), `area/*` (subsystem), `priority/*` (urgency), `triage/*` (review state), `lifecycle/*` (auto-close), `do-not-merge/*` (PR blockers), `security/*` (severity)
+  - For `area/*`: accuracy outweighs reuse. If no existing `area/*` truly fits the change, propose a new one via PR (extend `labels.yml` and the scope mapping in `pr-labeler.yaml`) — do not shoehorn the change into a wrong area. `area/uncategorized` is the auto-labeler fallback; treat it as a signal to pick a fit, create a new area, or correct the PR title
+  - PR titles: a Conventional Commits header (`type(scope): description`, types from [`contributing.md`](./docs/agents/contributing.md)) auto-applies `kind/*` and `area/*` via `.github/workflows/pr-labeler.yaml`. Append `!` (or add a `BREAKING CHANGE:` footer) to apply `kind/breaking-change`
+
 **Important rules:**
 - ✅ **ONLY read the file if the task matches the documented process scope** - do not read files for tasks that don't match their purpose
 - ✅ **ALWAYS read the file FIRST** before starting the task (when applicable)
diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 658dab74..6ff6eef1 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -57,6 +57,49 @@ git commit --signoff -m "fix(postgres): update operator to version 1.2.3"
 git commit --signoff -m "docs(contributing): add installation guide"
 ```
 
+## PR Title Auto-Labeling
+
+`.github/workflows/pr-labeler.yaml` parses the PR title on `opened`, `edited`, `reopened`, and `synchronize` events and applies labels additively (never removes). The title is expected to follow Conventional Commits — same format as commit messages above.
+
+**Type → `kind/*`:**
+
+| type      | label              |
+| --------- | ------------------ |
+| feat      | kind/feature       |
+| fix       | kind/bug           |
+| docs      | kind/documentation |
+| chore     | kind/cleanup       |
+| refactor  | kind/cleanup       |
+| style, perf, test, build, ci, revert | (no kind label)|
+
+**Scope → `area/*`** (full mapping in `.github/workflows/pr-labeler.yaml`):
+
+| scope (examples) | label |
+| --- | --- |
+| agents, ai | area/ai |
+| api, cozystack-api | area/api |
+| build | area/build |
+| ci | area/ci |
+| dashboard | area/dashboard |
+| postgres, mariadb, redis, etcd, kafka, clickhouse, postgres-operator, mariadb-operator | area/database |
+| extra | area/extra |
+| kubernetes | area/kubernetes |
+| monitoring, vlogs, vmstack, grafana, workloadmonitor | area/monitoring |
+| ingress, gateway, vpn, metallb, cilium, kube-ovn, cozy-proxy, … | area/networking |
+| platform, bundle, flux, fluxcd, cluster-api, talos, installer, cozyctl, cozystack-engine, cozy-lib | area/platform |
+| backport, release | area/release |
+| seaweedfs, bucket, linstor, velero, harbor, backups | area/storage |
+| tests, e2e | area/testing |
+| kubevirt, cdi, vmi, vm-import, virtual-machine, hami, gpu-operator | area/virtualization |
+
+**Special handling:**
+
+- `[Backport release-1.x]` prefix is stripped before parsing; `area/release` and `backport` labels are added.
+- Composite scope (`feat(platform, system, apps): …`) — each comma-separated part is mapped independently.
+- `!` after type or `BREAKING CHANGE:` footer in the body → `kind/breaking-change`.
+- Unmapped scope or non-conventional title → `area/uncategorized` (signals the PR needs manual area selection).
+- Bracket-style fallback (`[scope] description`) maps `scope` → `area/*` but cannot infer `kind/*`.
+
 ### AI Agent Attribution
 
 When an AI agent authors or materially assists with a commit, add an `Assisted-By:` trailer naming the model:

From 91188702a678c445bf42dafcbc3b697b1377b086 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 03:30:09 +0300
Subject: [PATCH 62/82] chore(ci): normalize hex color case in labels.yml

Address review feedback from gemini-code-assist on .github/labels.yml:242:
all hex color values use lowercase characters for consistency.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/labels.yml | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/.github/labels.yml b/.github/labels.yml
index 1369aa48..2ce04130 100644
--- a/.github/labels.yml
+++ b/.github/labels.yml
@@ -239,11 +239,11 @@
 # ──────────────────────────────────────────────
 
 - name: epic
-  color: 'A335EE'
+  color: 'a335ee'
   description: A large development increment that brings definite value to Cozystack users
 
 - name: community
-  color: '97458A'
+  color: '97458a'
   description: Community contributions are welcome in this issue
 
 - name: help wanted
@@ -263,7 +263,7 @@
   description: Requires resolving an issue in an upstream project
 
 - name: backport
-  color: 'FBCA04'
+  color: 'fbca04'
   description: Should change be backported on previous release
 
 - name: backport-previous
@@ -283,7 +283,7 @@
   description: Debugging in progress
 
 - name: sponsored
-  color: '00FF00'
+  color: '00ff00'
   description: Sponsored work
 
 - name: lgtm
@@ -291,7 +291,7 @@
   description: This PR has been approved by a maintainer
 
 - name: ok-to-test
-  color: '00FF00'
+  color: '00ff00'
   description: Indicates a non-member PR is safe to run CI on
 
 # ──────────────────────────────────────────────

From 31f4435eb0d02f905b6d0ba2f06c22034fc437d9 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 03:30:29 +0300
Subject: [PATCH 63/82] docs(agents): replace Unicode ellipsis with ASCII in
 contributing.md
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Address review feedback from gemini-code-assist on docs/agents/contributing.md:88:
… replaced with ... for compatibility across editors and tools.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 6ff6eef1..ac6aaf90 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -85,7 +85,7 @@ git commit --signoff -m "docs(contributing): add installation guide"
 | extra | area/extra |
 | kubernetes | area/kubernetes |
 | monitoring, vlogs, vmstack, grafana, workloadmonitor | area/monitoring |
-| ingress, gateway, vpn, metallb, cilium, kube-ovn, cozy-proxy, … | area/networking |
+| ingress, gateway, vpn, metallb, cilium, kube-ovn, cozy-proxy, ... | area/networking |
 | platform, bundle, flux, fluxcd, cluster-api, talos, installer, cozyctl, cozystack-engine, cozy-lib | area/platform |
 | backport, release | area/release |
 | seaweedfs, bucket, linstor, velero, harbor, backups | area/storage |
@@ -95,7 +95,7 @@ git commit --signoff -m "docs(contributing): add installation guide"
 **Special handling:**
 
 - `[Backport release-1.x]` prefix is stripped before parsing; `area/release` and `backport` labels are added.
-- Composite scope (`feat(platform, system, apps): …`) — each comma-separated part is mapped independently.
+- Composite scope (`feat(platform, system, apps): ...`) — each comma-separated part is mapped independently.
 - `!` after type or `BREAKING CHANGE:` footer in the body → `kind/breaking-change`.
 - Unmapped scope or non-conventional title → `area/uncategorized` (signals the PR needs manual area selection).
 - Bracket-style fallback (`[scope] description`) maps `scope` → `area/*` but cannot infer `kind/*`.

From 10b98ceb62308808aefbaa1ebdd4b1f07d713779 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 03:30:43 +0300
Subject: [PATCH 64/82] docs(agents): use full path .github/labels.yml in
 AGENTS.md

Address review feedback from gemini-code-assist on AGENTS.md:33:
expand bare labels.yml and pr-labeler.yaml to .github/labels.yml and
.github/workflows/pr-labeler.yaml for consistency with surrounding refs.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 AGENTS.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/AGENTS.md b/AGENTS.md
index 5298f6fd..1e9c4d15 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -30,7 +30,7 @@ working with the **Cozystack** project.
 - **Issue and PR labeling, triage** (e.g., "label this issue", "what label should I use", "triage this", "categorize")
   - Read: [`.github/labels.yml`](./.github/labels.yml)
   - Action: Use labels defined there. Conventions follow the Kubernetes scheme — `kind/*` (type), `area/*` (subsystem), `priority/*` (urgency), `triage/*` (review state), `lifecycle/*` (auto-close), `do-not-merge/*` (PR blockers), `security/*` (severity)
-  - For `area/*`: accuracy outweighs reuse. If no existing `area/*` truly fits the change, propose a new one via PR (extend `labels.yml` and the scope mapping in `pr-labeler.yaml`) — do not shoehorn the change into a wrong area. `area/uncategorized` is the auto-labeler fallback; treat it as a signal to pick a fit, create a new area, or correct the PR title
+  - For `area/*`: accuracy outweighs reuse. If no existing `area/*` truly fits the change, propose a new one via PR (extend `.github/labels.yml` and the scope mapping in `.github/workflows/pr-labeler.yaml`) — do not shoehorn the change into a wrong area. `area/uncategorized` is the auto-labeler fallback; treat it as a signal to pick a fit, create a new area, or correct the PR title
   - PR titles: a Conventional Commits header (`type(scope): description`, types from [`contributing.md`](./docs/agents/contributing.md)) auto-applies `kind/*` and `area/*` via `.github/workflows/pr-labeler.yaml`. Append `!` (or add a `BREAKING CHANGE:` footer) to apply `kind/breaking-change`
 
 **Important rules:**

From 738762994e10f878cab27a685a5dbf116ccc43fd Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 03:30:52 +0300
Subject: [PATCH 65/82] docs(agents): fix markdown table cell spacing in
 contributing.md

Address review feedback from gemini-code-assist on docs/agents/contributing.md:73:
add space before closing pipe in the type to kind mapping table for
consistency with other rows.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index ac6aaf90..d2ff94e2 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -70,7 +70,7 @@ git commit --signoff -m "docs(contributing): add installation guide"
 | docs      | kind/documentation |
 | chore     | kind/cleanup       |
 | refactor  | kind/cleanup       |
-| style, perf, test, build, ci, revert | (no kind label)|
+| style, perf, test, build, ci, revert | (no kind label) |
 
 **Scope → `area/*`** (full mapping in `.github/workflows/pr-labeler.yaml`):
 

From 104b3b3d2b1c25cc06a86669238cad3b744c86d3 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 20 Apr 2026 18:43:02 +0300
Subject: [PATCH 66/82] feat(operator): add per-package upgradeCRDs policy for
 HelmRelease

Add an opt-in UpgradeCRDs field to ComponentInstall that maps to
HelmRelease.Spec.Upgrade.CRDs, allowing a PackageSource component to
declare how Flux should handle CRDs from the chart's crds/ directory
on upgrade.

The helm-controller default on upgrade is Skip, which means new CRDs
added between chart versions never reach existing clusters and must be
applied manually. Setting upgradeCRDs: CreateReplace makes Flux apply
new CRDs declaratively with the chart.

Allowed values are restricted to Skip, Create, CreateReplace via a
kubebuilder enum marker. Empty / unset preserves the existing Flux
default, so all existing PackageSource resources keep working.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 api/v1alpha1/packagesource_types.go           |  10 ++
 .../cozystack.io_packagesources.yaml          |  13 ++
 internal/operator/package_reconciler.go       |  11 ++
 internal/operator/package_reconciler_test.go  | 138 ++++++++++++++++++
 4 files changed, 172 insertions(+)
 create mode 100644 internal/operator/package_reconciler_test.go

diff --git a/api/v1alpha1/packagesource_types.go b/api/v1alpha1/packagesource_types.go
index c0b943e8..75bf217b 100644
--- a/api/v1alpha1/packagesource_types.go
+++ b/api/v1alpha1/packagesource_types.go
@@ -132,6 +132,16 @@ type ComponentInstall struct {
 	// DependsOn is a list of component names that must be installed before this component
 	// +optional
 	DependsOn []string `json:"dependsOn,omitempty"`
+
+	// UpgradeCRDs controls how CRDs from the chart's crds/ directory are
+	// handled on HelmRelease upgrades. Maps to HelmRelease.Spec.Upgrade.CRDs.
+	// Empty string (default) preserves the helm-controller default (Skip).
+	// Use "CreateReplace" for operators that evolve their CRD set between
+	// versions. Warning: CreateReplace overwrites CRDs and may cause data
+	// loss if upstream drops fields from a CRD with live objects.
+	// +optional
+	// +kubebuilder:validation:Enum=Skip;Create;CreateReplace
+	UpgradeCRDs string `json:"upgradeCRDs,omitempty"`
 }
 
 // Component defines a single Helm release component within a package source
diff --git a/internal/crdinstall/manifests/cozystack.io_packagesources.yaml b/internal/crdinstall/manifests/cozystack.io_packagesources.yaml
index 0acfcdd2..1d0037df 100644
--- a/internal/crdinstall/manifests/cozystack.io_packagesources.yaml
+++ b/internal/crdinstall/manifests/cozystack.io_packagesources.yaml
@@ -118,6 +118,19 @@ spec:
                                   ReleaseName is the name of the HelmRelease resource that will be created
                                   If not specified, defaults to the component Name field
                                 type: string
+                              upgradeCRDs:
+                                description: |-
+                                  UpgradeCRDs controls how CRDs from the chart's crds/ directory are
+                                  handled on HelmRelease upgrades. Maps to HelmRelease.Spec.Upgrade.CRDs.
+                                  Empty string (default) preserves the helm-controller default (Skip).
+                                  Use "CreateReplace" for operators that evolve their CRD set between
+                                  versions. Warning: CreateReplace overwrites CRDs and may cause data
+                                  loss if upstream drops fields from a CRD with live objects.
+                                enum:
+                                - Skip
+                                - Create
+                                - CreateReplace
+                                type: string
                             type: object
                           libraries:
                             description: |-
diff --git a/internal/operator/package_reconciler.go b/internal/operator/package_reconciler.go
index 32e667e4..0e724e49 100644
--- a/internal/operator/package_reconciler.go
+++ b/internal/operator/package_reconciler.go
@@ -45,6 +45,16 @@ const (
 	SecretCozystackValues = "cozystack-values"
 )
 
+// parseCRDPolicy maps ComponentInstall.UpgradeCRDs to a helmv2.CRDsPolicy.
+// Empty / nil preserves the helm-controller default (Skip on upgrade);
+// the CRD enum marker restricts the string to Skip/Create/CreateReplace.
+func parseCRDPolicy(install *cozyv1alpha1.ComponentInstall) helmv2.CRDsPolicy {
+	if install == nil || install.UpgradeCRDs == "" {
+		return ""
+	}
+	return helmv2.CRDsPolicy(install.UpgradeCRDs)
+}
+
 // PackageReconciler reconciles Package resources
 type PackageReconciler struct {
 	client.Client
@@ -221,6 +231,7 @@ func (r *PackageReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ct
 					Remediation: &helmv2.UpgradeRemediation{
 						Retries: -1,
 					},
+					CRDs: parseCRDPolicy(component.Install),
 				},
 			},
 		}
diff --git a/internal/operator/package_reconciler_test.go b/internal/operator/package_reconciler_test.go
new file mode 100644
index 00000000..f0ee2c19
--- /dev/null
+++ b/internal/operator/package_reconciler_test.go
@@ -0,0 +1,138 @@
+/*
+Copyright 2025 The Cozystack Authors.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package operator
+
+import (
+	"encoding/json"
+	"os"
+	"path/filepath"
+	"testing"
+
+	cozyv1alpha1 "github.com/cozystack/cozystack/api/v1alpha1"
+	helmv2 "github.com/fluxcd/helm-controller/api/v2"
+	apiextensionsv1 "k8s.io/apiextensions-apiserver/pkg/apis/apiextensions/v1"
+	"sigs.k8s.io/yaml"
+)
+
+func TestParseCRDPolicy(t *testing.T) {
+	tests := []struct {
+		name    string
+		install *cozyv1alpha1.ComponentInstall
+		want    helmv2.CRDsPolicy
+	}{
+		{
+			name:    "nil install leaves flux default",
+			install: nil,
+			want:    "",
+		},
+		{
+			name:    "empty upgradeCRDs leaves flux default",
+			install: &cozyv1alpha1.ComponentInstall{},
+			want:    "",
+		},
+		{
+			name:    "Skip is passed through",
+			install: &cozyv1alpha1.ComponentInstall{UpgradeCRDs: "Skip"},
+			want:    helmv2.Skip,
+		},
+		{
+			name:    "Create is passed through",
+			install: &cozyv1alpha1.ComponentInstall{UpgradeCRDs: "Create"},
+			want:    helmv2.Create,
+		},
+		{
+			name:    "CreateReplace is passed through",
+			install: &cozyv1alpha1.ComponentInstall{UpgradeCRDs: "CreateReplace"},
+			want:    helmv2.CreateReplace,
+		},
+	}
+
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			got := parseCRDPolicy(tc.install)
+			if got != tc.want {
+				t.Errorf("parseCRDPolicy() = %q, want %q", got, tc.want)
+			}
+		})
+	}
+}
+
+// TestPackageSourceCRDHasUpgradeCRDsEnum guards the generated CRD schema: the
+// invalid-value case from the spec is enforced at the API server via a
+// kubebuilder enum marker, not in the reconciler. If someone drops the marker
+// and forgets to regenerate, this test catches it.
+func TestPackageSourceCRDHasUpgradeCRDsEnum(t *testing.T) {
+	path := filepath.Join("..", "crdinstall", "manifests", "cozystack.io_packagesources.yaml")
+	data, err := os.ReadFile(path)
+	if err != nil {
+		t.Fatalf("read %s: %v", path, err)
+	}
+
+	var crd apiextensionsv1.CustomResourceDefinition
+	if err := yaml.Unmarshal(data, &crd); err != nil {
+		t.Fatalf("unmarshal CRD: %v", err)
+	}
+
+	var field *apiextensionsv1.JSONSchemaProps
+	for i := range crd.Spec.Versions {
+		v := &crd.Spec.Versions[i]
+		if v.Schema == nil || v.Schema.OpenAPIV3Schema == nil {
+			continue
+		}
+		spec, ok := v.Schema.OpenAPIV3Schema.Properties["spec"]
+		if !ok {
+			continue
+		}
+		variants, ok := spec.Properties["variants"]
+		if !ok || variants.Items == nil || variants.Items.Schema == nil {
+			continue
+		}
+		components, ok := variants.Items.Schema.Properties["components"]
+		if !ok || components.Items == nil || components.Items.Schema == nil {
+			continue
+		}
+		install, ok := components.Items.Schema.Properties["install"]
+		if !ok {
+			continue
+		}
+		f, ok := install.Properties["upgradeCRDs"]
+		if !ok {
+			continue
+		}
+		field = &f
+		break
+	}
+
+	if field == nil {
+		t.Fatal("upgradeCRDs field not found in PackageSource CRD schema")
+	}
+
+	got := map[string]bool{}
+	for _, e := range field.Enum {
+		var s string
+		if err := json.Unmarshal(e.Raw, &s); err != nil {
+			t.Fatalf("unmarshal enum value %q: %v", e.Raw, err)
+		}
+		got[s] = true
+	}
+
+	for _, want := range []string{"Skip", "Create", "CreateReplace"} {
+		if !got[want] {
+			t.Errorf("enum value %q missing from upgradeCRDs; got %v", want, got)
+		}
+	}
+}

From d86bc7760a77c93c6fd837ef5e23c3df3b9b5898 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 20 Apr 2026 18:43:18 +0300
Subject: [PATCH 67/82] docs(agents): document PackageSource upgradeCRDs field

Describe when to set upgradeCRDs: CreateReplace (operators that evolve
their CRD set additively between versions) and the data-loss risk of
enabling it on operators that drop fields.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/overview.md | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/docs/agents/overview.md b/docs/agents/overview.md
index 35798961..ae0fd1c4 100644
--- a/docs/agents/overview.md
+++ b/docs/agents/overview.md
@@ -83,6 +83,14 @@ packages/<category>/<package-name>/
 - Reference PR numbers when available
 - Keep commits atomic and focused
 
+### PackageSource CRD upgrade policy
+
+Each component in a `PackageSource` may set `install.upgradeCRDs` to control how CRDs from the chart's `crds/` directory are handled on `HelmRelease` upgrades. Allowed values: `Skip` (default — helm-controller does not touch CRDs on upgrade), `Create` (create new CRDs only), `CreateReplace` (create new and overwrite existing).
+
+Set `upgradeCRDs: CreateReplace` for operators whose upstream regularly adds new CRDs between versions (etcd-operator, cnpg, kubevirt, kamaji). Without it, new CRDs from a chart bump do not land on existing clusters — only fresh installs get them.
+
+Do **not** set `CreateReplace` blindly: it overwrites every CRD in `crds/` and can cause silent data loss if upstream drops a field from a CRD that has live objects. Only enable it for operators whose schema evolution is additive-only. When in doubt, leave it unset and apply new CRDs manually.
+
 ### Documentation
 
 Documentation is organized as follows:

From f527ce683b28401c62cdc420188392b32367401f Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 20 Apr 2026 19:08:11 +0300
Subject: [PATCH 68/82] docs(agents): clarify that Scopes list is illustrative

The Scopes section was read as an exhaustive enumeration, which led to
review feedback flagging any scope outside the list as invalid. The
intent has always been that contributors pick the most specific scope
for the change and extend the list when a genuinely new area appears.
Reword the section accordingly and add operator as an example scope.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 docs/agents/contributing.md | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/docs/agents/contributing.md b/docs/agents/contributing.md
index 658dab74..9f3ff779 100644
--- a/docs/agents/contributing.md
+++ b/docs/agents/contributing.md
@@ -43,10 +43,11 @@ git commit --signoff -m "type(scope): brief description"
 
 **Types:** `feat`, `fix`, `docs`, `style`, `refactor`, `perf`, `test`, `build`, `ci`, `chore`
 
-**Scopes** (e.g., not exhaustive — use any scope that names the component you are touching):
-- System: `dashboard`, `platform`, `cilium`, `kube-ovn`, `linstor`, `fluxcd`, `cluster-api`
-- Apps: `postgres`, `mariadb`, `redis`, `kafka`, `clickhouse`, `virtual-machine`, `kubernetes`
-- Other: `api`, `hack`, `tests`, `ci`, `docs`, `agents`, `maintenance`
+**Scopes** (examples — not an exhaustive list; pick the most specific scope that describes the change, and introduce a new one if a genuinely new area needs its own):
+
+- System, e.g.: `dashboard`, `platform`, `operator`, `cilium`, `kube-ovn`, `linstor`, `fluxcd`, `cluster-api`
+- Apps, e.g.: `postgres`, `mariadb`, `redis`, `kafka`, `clickhouse`, `virtual-machine`, `kubernetes`
+- Other, e.g.: `api`, `hack`, `tests`, `ci`, `docs`, `agents`, `maintenance`
 
 Breaking changes: append `!` after type/scope (`feat(api)!: ...`) or add a `BREAKING CHANGE:` footer.
 

From 32ae993d3ebb51a34ece2b6bbe8e6626c72ac300 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 20 Apr 2026 19:13:06 +0300
Subject: [PATCH 69/82] docs(maintenance): mirror illustrative-scopes wording
 in PR template

Match the wording adopted in docs/agents/contributing.md so that human
contributors and AI agents see the same guidance in both places.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/PULL_REQUEST_TEMPLATE.md | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md
index b9475846..044dd6a0 100644
--- a/.github/PULL_REQUEST_TEMPLATE.md
+++ b/.github/PULL_REQUEST_TEMPLATE.md
@@ -1,9 +1,10 @@
 <!-- Thank you for making a contribution! Here are some tips for you:
 - Use Conventional Commits for the PR title: `type(scope): description`
   - Types: feat, fix, docs, style, refactor, perf, test, build, ci, chore
-  - Scopes for system components: dashboard, platform, cilium, kube-ovn, linstor, fluxcd, cluster-api
-  - Scopes for managed apps: postgres, mariadb, redis, kafka, clickhouse, virtual-machine, kubernetes
-  - Scopes for development and maintenance: api, hack, tests, ci, docs, maintenance
+  - Scopes are not an exhaustive list — pick the most specific scope for the change and extend the list when a genuinely new area appears. Examples:
+    - System components: dashboard, platform, operator, cilium, kube-ovn, linstor, fluxcd, cluster-api
+    - Managed apps: postgres, mariadb, redis, kafka, clickhouse, virtual-machine, kubernetes
+    - Development and maintenance: api, hack, tests, ci, docs, maintenance
   - Breaking changes: append `!` after type/scope (`feat(api)!: ...`) or add a `BREAKING CHANGE:` footer
 - If it's a work in progress, consider creating this PR as a draft.
 - Don't hesistate to ask for opinion and review in the community chats, even if it's still a draft.

From 1baadd75a72ec75921df438671b5e186aed063e2 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 13:07:04 +0300
Subject: [PATCH 70/82] fix(ci): guard pr.labels access in pr-labeler workflow

Address review feedback from myasnikovdaniil on .github/workflows/pr-labeler.yaml:129:
defensive (pr.labels || []) avoids TypeError if the webhook payload
arrives without the labels field on edge cases like stripped edited
events.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/workflows/pr-labeler.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/pr-labeler.yaml b/.github/workflows/pr-labeler.yaml
index d882bc50..ca3023fd 100644
--- a/.github/workflows/pr-labeler.yaml
+++ b/.github/workflows/pr-labeler.yaml
@@ -126,7 +126,7 @@ jobs:
             const pr = context.payload.pull_request;
             const title = pr.title || '';
             const body  = pr.body  || '';
-            const existing = new Set(pr.labels.map(l => l.name));
+            const existing = new Set((pr.labels || []).map(l => l.name));
             const toAdd = new Set();
 
             // 1. Strip "[Backport release-1.x]" prefix if present.

From 86a1e811ccde200d5206b98df70c4cc29a891ccc Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 13:07:21 +0300
Subject: [PATCH 71/82] fix(ci): accept hyphenated BREAKING-CHANGE footer in
 pr-labeler

Address review feedback from myasnikovdaniil on .github/workflows/pr-labeler.yaml:155:
Conventional Commits 1.0 spec item 16 treats BREAKING CHANGE: and
BREAKING-CHANGE: as synonymous footers. The hyphen form was silently
ignored before, so PRs that use it would miss kind/breaking-change.

https://www.conventionalcommits.org/en/v1.0.0/#specification

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/workflows/pr-labeler.yaml | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/pr-labeler.yaml b/.github/workflows/pr-labeler.yaml
index ca3023fd..5a6564a9 100644
--- a/.github/workflows/pr-labeler.yaml
+++ b/.github/workflows/pr-labeler.yaml
@@ -151,8 +151,9 @@ jobs:
               scopeStr = bracket[1];
             }
 
-            // 4. Detect BREAKING CHANGE: footer in body.
-            if (/^BREAKING CHANGE:/m.test(body)) {
+            // 4. Detect BREAKING CHANGE: or BREAKING-CHANGE: footer in body.
+            // Conventional Commits 1.0 spec item 16 treats them as synonymous.
+            if (/^BREAKING[ -]CHANGE:/m.test(body)) {
               breaking = true;
             }
 

From 79744099f63cfa3498362b15f9d2d30024602d0c Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 13:07:39 +0300
Subject: [PATCH 72/82] chore(ci): warn on unmapped type or scope in pr-labeler

Address review feedback from myasnikovdaniil on .github/workflows/pr-labeler.yaml:160,173:
emit core.warning when a Conventional Commits type has no kind/*
mapping or a scope has no area/* mapping. Without the warning, typos
(e.g., "hotfix" instead of "fix") and recurring new scopes silently
fall through to area/uncategorized, masking that the mapping has
drifted.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 .github/workflows/pr-labeler.yaml | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/pr-labeler.yaml b/.github/workflows/pr-labeler.yaml
index 5a6564a9..6e6f3017 100644
--- a/.github/workflows/pr-labeler.yaml
+++ b/.github/workflows/pr-labeler.yaml
@@ -158,8 +158,12 @@ jobs:
             }
 
             // 5. Apply kind/* from type.
-            if (type && typeToKind[type]) {
-              toAdd.add(typeToKind[type]);
+            if (type) {
+              if (typeToKind[type]) {
+                toAdd.add(typeToKind[type]);
+              } else {
+                core.warning(`type "${type}" has no kind/* mapping — typo or new type? See .github/workflows/pr-labeler.yaml typeToKind`);
+              }
             }
 
             // 6. Apply area/* from scope. Composite scopes split on comma.
@@ -170,6 +174,8 @@ jobs:
             for (const s of scopes) {
               if (scopeToArea[s]) {
                 toAdd.add(scopeToArea[s]);
+              } else {
+                core.warning(`scope "${s}" has no area/* mapping — consider extending scopeToArea in .github/workflows/pr-labeler.yaml if it recurs`);
               }
             }
 

From 37ecd7b3af9f96ef8cbcb097b8b88a4ffbcef548 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 13:50:08 +0300
Subject: [PATCH 73/82] feat(kubernetes): make wait-for-kubeconfig image
 overridable for air-gapped registries

Operators in air-gapped or rate-limited environments cannot reach
docker.io and the bundled busybox digest pin gives them no escape
hatch. Add an optional images.waitForKubeconfig chart value that, when
set, replaces the helper's image reference with any registry path
kubelet can pull. Empty value falls back to images/busybox.tag, so the
prior digest-pinned default is preserved.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 api/apps/v1alpha1/kubernetes/types.go         |  9 ++++
 packages/apps/kubernetes/README.md            | 52 ++++++++++---------
 .../apps/kubernetes/templates/_helpers.tpl    | 12 +++--
 .../tests/admin_kubeconfig_wait_test.yaml     | 23 ++++++++
 packages/apps/kubernetes/values.schema.json   | 12 +++++
 packages/apps/kubernetes/values.yaml          |  7 +++
 .../kubernetes-rd/cozyrds/kubernetes.yaml     |  4 +-
 7 files changed, 87 insertions(+), 32 deletions(-)

diff --git a/api/apps/v1alpha1/kubernetes/types.go b/api/apps/v1alpha1/kubernetes/types.go
index 774fcdc5..ea4a2b64 100644
--- a/api/apps/v1alpha1/kubernetes/types.go
+++ b/api/apps/v1alpha1/kubernetes/types.go
@@ -36,6 +36,9 @@ type ConfigSpec struct {
 	// Kubernetes control-plane configuration.
 	// +kubebuilder:default:={}
 	ControlPlane ControlPlane `json:"controlPlane"`
+	// Optional image overrides for air-gapped or rate-limited registries.
+	// +kubebuilder:default:={}
+	Images Images `json:"images"`
 }
 
 type APIServer struct {
@@ -157,6 +160,12 @@ type GatewayAPIAddon struct {
 	Enabled bool `json:"enabled"`
 }
 
+type Images struct {
+	// Image used by the wait-for-kubeconfig init container. Empty falls back to images/busybox.tag.
+	// +kubebuilder:default:=""
+	WaitForKubeconfig string `json:"waitForKubeconfig,omitempty"`
+}
+
 type IngressNginxAddon struct {
 	// Enable the controller (requires nodes labeled `ingress-nginx`).
 	// +kubebuilder:default:=false
diff --git a/packages/apps/kubernetes/README.md b/packages/apps/kubernetes/README.md
index d62d3a39..ecbbc31b 100644
--- a/packages/apps/kubernetes/README.md
+++ b/packages/apps/kubernetes/README.md
@@ -145,31 +145,33 @@ See the reference for components utilized in this service:
 
 ### Kubernetes Control Plane Configuration
 
-| Name                                                | Description                                      | Type       | Value   |
-| --------------------------------------------------- | ------------------------------------------------ | ---------- | ------- |
-| `controlPlane`                                      | Kubernetes control-plane configuration.          | `object`   | `{}`    |
-| `controlPlane.replicas`                             | Number of control-plane replicas.                | `int`      | `2`     |
-| `controlPlane.apiServer`                            | API Server configuration.                        | `object`   | `{}`    |
-| `controlPlane.apiServer.resources`                  | CPU and memory resources for API Server.         | `object`   | `{}`    |
-| `controlPlane.apiServer.resources.cpu`              | CPU available.                                   | `quantity` | `""`    |
-| `controlPlane.apiServer.resources.memory`           | Memory (RAM) available.                          | `quantity` | `""`    |
-| `controlPlane.apiServer.resourcesPreset`            | Preset if `resources` omitted.                   | `string`   | `large` |
-| `controlPlane.controllerManager`                    | Controller Manager configuration.                | `object`   | `{}`    |
-| `controlPlane.controllerManager.resources`          | CPU and memory resources for Controller Manager. | `object`   | `{}`    |
-| `controlPlane.controllerManager.resources.cpu`      | CPU available.                                   | `quantity` | `""`    |
-| `controlPlane.controllerManager.resources.memory`   | Memory (RAM) available.                          | `quantity` | `""`    |
-| `controlPlane.controllerManager.resourcesPreset`    | Preset if `resources` omitted.                   | `string`   | `micro` |
-| `controlPlane.scheduler`                            | Scheduler configuration.                         | `object`   | `{}`    |
-| `controlPlane.scheduler.resources`                  | CPU and memory resources for Scheduler.          | `object`   | `{}`    |
-| `controlPlane.scheduler.resources.cpu`              | CPU available.                                   | `quantity` | `""`    |
-| `controlPlane.scheduler.resources.memory`           | Memory (RAM) available.                          | `quantity` | `""`    |
-| `controlPlane.scheduler.resourcesPreset`            | Preset if `resources` omitted.                   | `string`   | `micro` |
-| `controlPlane.konnectivity`                         | Konnectivity configuration.                      | `object`   | `{}`    |
-| `controlPlane.konnectivity.server`                  | Konnectivity Server configuration.               | `object`   | `{}`    |
-| `controlPlane.konnectivity.server.resources`        | CPU and memory resources for Konnectivity.       | `object`   | `{}`    |
-| `controlPlane.konnectivity.server.resources.cpu`    | CPU available.                                   | `quantity` | `""`    |
-| `controlPlane.konnectivity.server.resources.memory` | Memory (RAM) available.                          | `quantity` | `""`    |
-| `controlPlane.konnectivity.server.resourcesPreset`  | Preset if `resources` omitted.                   | `string`   | `micro` |
+| Name                                                | Description                                                                                   | Type       | Value   |
+| --------------------------------------------------- | --------------------------------------------------------------------------------------------- | ---------- | ------- |
+| `controlPlane`                                      | Kubernetes control-plane configuration.                                                       | `object`   | `{}`    |
+| `controlPlane.replicas`                             | Number of control-plane replicas.                                                             | `int`      | `2`     |
+| `controlPlane.apiServer`                            | API Server configuration.                                                                     | `object`   | `{}`    |
+| `controlPlane.apiServer.resources`                  | CPU and memory resources for API Server.                                                      | `object`   | `{}`    |
+| `controlPlane.apiServer.resources.cpu`              | CPU available.                                                                                | `quantity` | `""`    |
+| `controlPlane.apiServer.resources.memory`           | Memory (RAM) available.                                                                       | `quantity` | `""`    |
+| `controlPlane.apiServer.resourcesPreset`            | Preset if `resources` omitted.                                                                | `string`   | `large` |
+| `controlPlane.controllerManager`                    | Controller Manager configuration.                                                             | `object`   | `{}`    |
+| `controlPlane.controllerManager.resources`          | CPU and memory resources for Controller Manager.                                              | `object`   | `{}`    |
+| `controlPlane.controllerManager.resources.cpu`      | CPU available.                                                                                | `quantity` | `""`    |
+| `controlPlane.controllerManager.resources.memory`   | Memory (RAM) available.                                                                       | `quantity` | `""`    |
+| `controlPlane.controllerManager.resourcesPreset`    | Preset if `resources` omitted.                                                                | `string`   | `micro` |
+| `controlPlane.scheduler`                            | Scheduler configuration.                                                                      | `object`   | `{}`    |
+| `controlPlane.scheduler.resources`                  | CPU and memory resources for Scheduler.                                                       | `object`   | `{}`    |
+| `controlPlane.scheduler.resources.cpu`              | CPU available.                                                                                | `quantity` | `""`    |
+| `controlPlane.scheduler.resources.memory`           | Memory (RAM) available.                                                                       | `quantity` | `""`    |
+| `controlPlane.scheduler.resourcesPreset`            | Preset if `resources` omitted.                                                                | `string`   | `micro` |
+| `controlPlane.konnectivity`                         | Konnectivity configuration.                                                                   | `object`   | `{}`    |
+| `controlPlane.konnectivity.server`                  | Konnectivity Server configuration.                                                            | `object`   | `{}`    |
+| `controlPlane.konnectivity.server.resources`        | CPU and memory resources for Konnectivity.                                                    | `object`   | `{}`    |
+| `controlPlane.konnectivity.server.resources.cpu`    | CPU available.                                                                                | `quantity` | `""`    |
+| `controlPlane.konnectivity.server.resources.memory` | Memory (RAM) available.                                                                       | `quantity` | `""`    |
+| `controlPlane.konnectivity.server.resourcesPreset`  | Preset if `resources` omitted.                                                                | `string`   | `micro` |
+| `images`                                            | Optional image overrides for air-gapped or rate-limited registries.                           | `object`   | `{}`    |
+| `images.waitForKubeconfig`                          | Image used by the wait-for-kubeconfig init container. Empty falls back to images/busybox.tag. | `string`   | `""`    |
 
 
 ## Parameter examples and reference
diff --git a/packages/apps/kubernetes/templates/_helpers.tpl b/packages/apps/kubernetes/templates/_helpers.tpl
index e6ad9dcf..40a8ae7a 100644
--- a/packages/apps/kubernetes/templates/_helpers.tpl
+++ b/packages/apps/kubernetes/templates/_helpers.tpl
@@ -65,18 +65,20 @@ release.cozystack.io/helm-install-timeout annotation) so the
 CrashLoopBackOff surfaces before flux remediation fires and uninstalls
 the Cluster CR.
 
-The pinned busybox image in images/busybox.tag points directly at
+The default image lives in images/busybox.tag and points directly at
 docker.io by digest (not mirrored to ghcr.io like the other .tag files
-here): the payload is a one-shot sh loop, the digest pin makes the
-pull immutable, and the cost of maintaining a private mirror of a tiny
-upstream image that does not move often is not worth it.
+here): the payload is a one-shot sh loop and the digest pin makes the
+pull immutable. Operators in air-gapped or rate-limited environments
+can override it via .Values.images.waitForKubeconfig (any registry
+reference kubelet can pull). When the value is empty the chart falls
+back to the bundled digest pin, preserving the prior default.
 
 Call site owns the surrounding volumes block; the kubeconfig volume
 must exist on the pod and mount at /etc/kubernetes/kubeconfig.
 */}}
 {{- define "kubernetes.waitForAdminKubeconfig" -}}
 - name: wait-for-kubeconfig
-  image: "{{ .Files.Get "images/busybox.tag" | trim }}"
+  image: "{{ default (.Files.Get "images/busybox.tag" | trim) .Values.images.waitForKubeconfig }}"
   command:
   - sh
   - -c
diff --git a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
index e7c03bd1..507e00af 100644
--- a/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
+++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml
@@ -95,6 +95,29 @@ tests:
             mountPath: /etc/kubernetes/kubeconfig
             readOnly: true
 
+  - it: wait-for-kubeconfig defaults to bundled busybox digest pin
+    template: templates/cluster-autoscaler/deployment.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
+    asserts:
+      - matchRegex:
+          path: spec.template.spec.initContainers[0].image
+          pattern: '^docker\.io/library/busybox:[^@]+@sha256:[0-9a-f]{64}$'
+
+  - it: wait-for-kubeconfig honours images.waitForKubeconfig override
+    template: templates/cluster-autoscaler/deployment.yaml
+    documentSelector:
+      path: kind
+      value: Deployment
+    set:
+      images:
+        waitForKubeconfig: "registry.example.test/library/busybox:1.37.0@sha256:0000000000000000000000000000000000000000000000000000000000000000"
+    asserts:
+      - equal:
+          path: spec.template.spec.initContainers[0].image
+          value: "registry.example.test/library/busybox:1.37.0@sha256:0000000000000000000000000000000000000000000000000000000000000000"
+
   - it: cluster.yaml renders and wires dataStoreName when tenant has etcd
     template: templates/cluster.yaml
     documentSelector:
diff --git a/packages/apps/kubernetes/values.schema.json b/packages/apps/kubernetes/values.schema.json
index 1beeff9c..81e9e2b8 100644
--- a/packages/apps/kubernetes/values.schema.json
+++ b/packages/apps/kubernetes/values.schema.json
@@ -630,6 +630,18 @@
           }
         }
       }
+    },
+    "images": {
+      "description": "Optional image overrides for air-gapped or rate-limited registries.",
+      "type": "object",
+      "default": {},
+      "properties": {
+        "waitForKubeconfig": {
+          "description": "Image used by the wait-for-kubeconfig init container. Empty falls back to images/busybox.tag.",
+          "type": "string",
+          "default": ""
+        }
+      }
     }
   }
 }
diff --git a/packages/apps/kubernetes/values.yaml b/packages/apps/kubernetes/values.yaml
index a67b5d69..68ba07b0 100644
--- a/packages/apps/kubernetes/values.yaml
+++ b/packages/apps/kubernetes/values.yaml
@@ -197,3 +197,10 @@ controlPlane:
     server:
       resources: {}
       resourcesPreset: "micro"
+
+## @typedef {struct} Images - Optional image overrides for chart-internal helpers.
+## @field {string} [waitForKubeconfig] - Image used by the wait-for-kubeconfig init container. Empty falls back to images/busybox.tag.
+
+## @param {Images} images - Optional image overrides for air-gapped or rate-limited registries.
+images:
+  waitForKubeconfig: ""
diff --git a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
index aaa01247..1631c86e 100644
--- a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
+++ b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml
@@ -22,7 +22,7 @@ spec:
     singular: kubernetes
     plural: kuberneteses
     openAPISchema: |-
-      {"title":"Chart Values","type":"object","properties":{"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":"replicated"},"nodeGroups":{"description":"Worker nodes configuration map.","type":"object","default":{"md0":{"ephemeralStorage":"20Gi","gpus":[],"instanceType":"u1.medium","maxReplicas":10,"minReplicas":0,"resources":{},"roles":["ingress-nginx"]}},"additionalProperties":{"type":"object","required":["ephemeralStorage","instanceType","maxReplicas","minReplicas","resources"],"properties":{"ephemeralStorage":{"description":"Ephemeral storage size.","default":"20Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"gpus":{"description":"List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).","type":"array","items":{"type":"object","required":["name"],"properties":{"name":{"description":"Name of GPU, such as \"nvidia.com/AD102GL_L40S\".","type":"string"}}}},"instanceType":{"description":"Virtual machine instance type.","type":"string","default":"u1.medium"},"maxReplicas":{"description":"Maximum number of replicas.","type":"integer","default":10},"minReplicas":{"description":"Minimum number of replicas.","type":"integer","default":0},"resources":{"description":"CPU and memory resources for each worker node.","type":"object","properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"roles":{"description":"List of node roles.","type":"array","items":{"type":"string"}}}}},"version":{"description":"Kubernetes major.minor version to deploy","type":"string","default":"v1.35","enum":["v1.35","v1.34","v1.33","v1.32","v1.31","v1.30"]},"host":{"description":"External hostname for Kubernetes cluster. Defaults to `<cluster-name>.<tenant-host>` if empty.","type":"string","default":""},"addons":{"description":"Cluster addons configuration.","type":"object","default":{},"required":["certManager","cilium","coredns","fluxcd","gatewayAPI","gpuOperator","ingressNginx","monitoringAgents","velero","verticalPodAutoscaler"],"properties":{"certManager":{"description":"Cert-manager addon.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable cert-manager.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"cilium":{"description":"Cilium CNI plugin.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"coredns":{"description":"CoreDNS addon.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"fluxcd":{"description":"FluxCD GitOps operator.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable FluxCD.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"gatewayAPI":{"description":"Gateway API addon.","type":"object","default":{},"required":["enabled"],"properties":{"enabled":{"description":"Enable Gateway API.","type":"boolean","default":false}}},"gpuOperator":{"description":"NVIDIA GPU Operator.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable GPU Operator.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"ingressNginx":{"description":"Ingress-NGINX controller.","type":"object","default":{},"required":["enabled","exposeMethod","valuesOverride"],"properties":{"enabled":{"description":"Enable the controller (requires nodes labeled `ingress-nginx`).","type":"boolean","default":false},"exposeMethod":{"description":"Method to expose the controller. Allowed values: `Proxied`, `LoadBalancer`.","type":"string","default":"Proxied","enum":["Proxied","LoadBalancer"]},"hosts":{"description":"Domains routed to this tenant cluster when `exposeMethod` is `Proxied`.","type":"array","default":[],"items":{"type":"string"}},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"monitoringAgents":{"description":"Monitoring agents.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable monitoring agents.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"velero":{"description":"Velero backup/restore addon.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable Velero.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"verticalPodAutoscaler":{"description":"Vertical Pod Autoscaler.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}}}},"controlPlane":{"description":"Kubernetes control-plane configuration.","type":"object","default":{},"required":["apiServer","controllerManager","konnectivity","replicas","scheduler"],"properties":{"apiServer":{"description":"API Server configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for API Server.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"large","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}},"controllerManager":{"description":"Controller Manager configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Controller Manager.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}},"konnectivity":{"description":"Konnectivity configuration.","type":"object","default":{},"required":["server"],"properties":{"server":{"description":"Konnectivity Server configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Konnectivity.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}}}},"replicas":{"description":"Number of control-plane replicas.","type":"integer","default":2},"scheduler":{"description":"Scheduler configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Scheduler.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}}}}}}
+      {"title":"Chart Values","type":"object","properties":{"storageClass":{"description":"StorageClass used to store the data.","type":"string","default":"replicated"},"nodeGroups":{"description":"Worker nodes configuration map.","type":"object","default":{"md0":{"ephemeralStorage":"20Gi","gpus":[],"instanceType":"u1.medium","maxReplicas":10,"minReplicas":0,"resources":{},"roles":["ingress-nginx"]}},"additionalProperties":{"type":"object","required":["ephemeralStorage","instanceType","maxReplicas","minReplicas","resources"],"properties":{"ephemeralStorage":{"description":"Ephemeral storage size.","default":"20Gi","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"gpus":{"description":"List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).","type":"array","items":{"type":"object","required":["name"],"properties":{"name":{"description":"Name of GPU, such as \"nvidia.com/AD102GL_L40S\".","type":"string"}}}},"instanceType":{"description":"Virtual machine instance type.","type":"string","default":"u1.medium"},"maxReplicas":{"description":"Maximum number of replicas.","type":"integer","default":10},"minReplicas":{"description":"Minimum number of replicas.","type":"integer","default":0},"resources":{"description":"CPU and memory resources for each worker node.","type":"object","properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"roles":{"description":"List of node roles.","type":"array","items":{"type":"string"}}}}},"version":{"description":"Kubernetes major.minor version to deploy","type":"string","default":"v1.35","enum":["v1.35","v1.34","v1.33","v1.32","v1.31","v1.30"]},"host":{"description":"External hostname for Kubernetes cluster. Defaults to `<cluster-name>.<tenant-host>` if empty.","type":"string","default":""},"addons":{"description":"Cluster addons configuration.","type":"object","default":{},"required":["certManager","cilium","coredns","fluxcd","gatewayAPI","gpuOperator","ingressNginx","monitoringAgents","velero","verticalPodAutoscaler"],"properties":{"certManager":{"description":"Cert-manager addon.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable cert-manager.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"cilium":{"description":"Cilium CNI plugin.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"coredns":{"description":"CoreDNS addon.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"fluxcd":{"description":"FluxCD GitOps operator.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable FluxCD.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"gatewayAPI":{"description":"Gateway API addon.","type":"object","default":{},"required":["enabled"],"properties":{"enabled":{"description":"Enable Gateway API.","type":"boolean","default":false}}},"gpuOperator":{"description":"NVIDIA GPU Operator.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable GPU Operator.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"ingressNginx":{"description":"Ingress-NGINX controller.","type":"object","default":{},"required":["enabled","exposeMethod","valuesOverride"],"properties":{"enabled":{"description":"Enable the controller (requires nodes labeled `ingress-nginx`).","type":"boolean","default":false},"exposeMethod":{"description":"Method to expose the controller. Allowed values: `Proxied`, `LoadBalancer`.","type":"string","default":"Proxied","enum":["Proxied","LoadBalancer"]},"hosts":{"description":"Domains routed to this tenant cluster when `exposeMethod` is `Proxied`.","type":"array","default":[],"items":{"type":"string"}},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"monitoringAgents":{"description":"Monitoring agents.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable monitoring agents.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"velero":{"description":"Velero backup/restore addon.","type":"object","default":{},"required":["enabled","valuesOverride"],"properties":{"enabled":{"description":"Enable Velero.","type":"boolean","default":false},"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}},"verticalPodAutoscaler":{"description":"Vertical Pod Autoscaler.","type":"object","default":{},"required":["valuesOverride"],"properties":{"valuesOverride":{"description":"Custom Helm values overrides.","type":"object","default":{},"x-kubernetes-preserve-unknown-fields":true}}}}},"controlPlane":{"description":"Kubernetes control-plane configuration.","type":"object","default":{},"required":["apiServer","controllerManager","konnectivity","replicas","scheduler"],"properties":{"apiServer":{"description":"API Server configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for API Server.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"large","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}},"controllerManager":{"description":"Controller Manager configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Controller Manager.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}},"konnectivity":{"description":"Konnectivity configuration.","type":"object","default":{},"required":["server"],"properties":{"server":{"description":"Konnectivity Server configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Konnectivity.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}}}},"replicas":{"description":"Number of control-plane replicas.","type":"integer","default":2},"scheduler":{"description":"Scheduler configuration.","type":"object","default":{},"required":["resources","resourcesPreset"],"properties":{"resources":{"description":"CPU and memory resources for Scheduler.","type":"object","default":{},"properties":{"cpu":{"description":"CPU available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Memory (RAM) available.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"resourcesPreset":{"description":"Preset if `resources` omitted.","type":"string","default":"micro","enum":["nano","micro","small","medium","large","xlarge","2xlarge"]}}}}},"images":{"description":"Optional image overrides for air-gapped or rate-limited registries.","type":"object","default":{},"properties":{"waitForKubeconfig":{"description":"Image used by the wait-for-kubeconfig init container. Empty falls back to images/busybox.tag.","type":"string","default":""}}}}}
   release:
     prefix: kubernetes-
     labels:
@@ -40,7 +40,7 @@ spec:
     tags:
       - compute
     icon: <svg width="144" height="144" viewBox="0 0 144 144" fill="none" xmlns="http://www.w3.org/2000/svg">
<rect width="144" height="144" rx="24" fill="url(#paint0_linear_681_2845)"/>
<path d="M71.9968 19C70.3039 19.0002 68.9312 20.5322 68.9314 22.4221C68.9314 22.4511 68.9373 22.4788 68.9379 22.5076C68.9354 22.7644 68.9231 23.0737 68.9314 23.2973C68.9717 24.3873 69.2082 25.2216 69.3506 26.2258C69.6084 28.3752 69.8245 30.1569 69.6912 31.813C69.5615 32.4375 69.1037 33.0086 68.6956 33.4056L68.6235 34.7086C66.7839 34.8617 64.9319 35.1421 63.0821 35.5641C55.1226 37.3798 48.2695 41.4991 43.052 47.0609C42.7134 46.8288 42.1211 46.4019 41.945 46.2712C41.3977 46.3454 40.8445 46.5151 40.1241 46.0935C38.7522 45.1657 37.5028 43.8851 35.991 42.3424C35.2982 41.6044 34.7966 40.9018 33.9735 40.1904C33.7866 40.0289 33.5014 39.8104 33.2923 39.6442C32.6489 39.1288 31.89 38.86 31.157 38.8348C30.2147 38.8024 29.3075 39.1725 28.7138 39.9206C27.6584 41.2506 27.9963 43.2833 29.4671 44.4614C29.482 44.4734 29.4979 44.4827 29.5129 44.4943C29.715 44.6589 29.9625 44.8698 30.1483 45.0076C31.0217 45.6555 31.8195 45.9872 32.6897 46.5015C34.5231 47.6391 36.043 48.5823 37.2486 49.7196C37.7194 50.2237 37.8016 51.1122 37.8643 51.4964L38.8468 52.3782C33.5872 60.3308 31.153 70.1539 32.5915 80.1627L31.3077 80.5378C30.9693 80.9768 30.4912 81.6676 29.9911 81.8738C28.4138 82.3729 26.6387 82.5562 24.4956 82.7819C23.4894 82.866 22.6213 82.8158 21.5546 83.0188C21.3198 83.0635 20.9927 83.1491 20.7358 83.2097C20.7269 83.2116 20.7186 83.2142 20.7096 83.2162C20.6956 83.2195 20.6772 83.2263 20.6638 83.2294C18.857 83.668 17.6963 85.3365 18.0699 86.9805C18.4437 88.6248 20.2086 89.6248 22.0262 89.2312C22.0393 89.2282 22.0584 89.2277 22.072 89.2246C22.0926 89.2199 22.1106 89.2099 22.131 89.2049C22.3844 89.149 22.7019 89.0868 22.9236 89.0272C23.9723 88.7451 24.7318 88.3306 25.6746 87.9677C27.7029 87.2368 29.3828 86.6262 31.0195 86.3883C31.703 86.3345 32.4232 86.812 32.7814 87.0134L34.1177 86.7831C37.1926 96.3613 43.6366 104.103 51.7963 108.961L51.2396 110.303C51.4403 110.824 51.6616 111.53 51.5121 112.045C50.9171 113.595 49.898 115.231 48.7374 117.055C48.1755 117.898 47.6004 118.552 47.0934 119.516C46.972 119.747 46.8175 120.102 46.7004 120.346C45.9125 122.039 46.4904 123.99 48.0038 124.722C49.5268 125.459 51.4171 124.682 52.2352 122.985C52.2364 122.982 52.2406 122.98 52.2417 122.978C52.2426 122.976 52.2409 122.973 52.2417 122.971C52.3582 122.731 52.5233 122.415 52.6216 122.188C53.056 121.189 53.2005 120.332 53.5059 119.365C54.317 117.318 54.7626 115.17 55.8791 113.832C56.1849 113.466 56.6833 113.325 57.2001 113.186L57.8944 111.922C65.008 114.665 72.9705 115.402 80.9245 113.587C82.7391 113.173 84.4908 112.637 86.1843 111.994C86.3794 112.342 86.742 113.011 86.8393 113.179C87.3644 113.351 87.9377 113.439 88.4047 114.133C89.2401 115.567 89.8114 117.263 90.5073 119.312C90.8128 120.279 90.9638 121.136 91.3981 122.136C91.4971 122.363 91.6614 122.684 91.778 122.925C92.5944 124.628 94.4907 125.407 96.0159 124.669C97.5292 123.937 98.1077 121.986 97.3194 120.293C97.2023 120.049 97.0412 119.695 96.9198 119.464C96.4127 118.499 95.8377 117.852 95.2758 117.009C94.1152 115.185 93.1526 113.67 92.5575 112.12C92.3087 111.32 92.5995 110.823 92.7933 110.303C92.6772 110.17 92.4288 109.414 92.2824 109.059C100.762 104.029 107.017 95.9985 109.955 86.7239C110.351 86.7865 111.041 86.9091 111.265 86.9542C111.726 86.6487 112.149 86.2501 112.981 86.3159C114.617 86.5537 116.297 87.1645 118.326 87.8953C119.268 88.2581 120.028 88.6793 121.077 88.9614C121.298 89.021 121.616 89.0766 121.869 89.1325C121.889 89.1375 121.908 89.1475 121.928 89.1522C121.942 89.1553 121.961 89.1558 121.974 89.1588C123.792 89.552 125.557 88.5526 125.93 86.9081C126.303 85.2641 125.143 83.5952 123.336 83.157C123.074 83.097 122.701 82.995 122.446 82.9465C121.379 82.7435 120.511 82.7935 119.505 82.7095C117.361 82.4839 115.586 82.3004 114.009 81.8014C113.366 81.5507 112.908 80.7819 112.686 80.4655L111.448 80.1035C112.09 75.438 111.917 70.5825 110.806 65.7243C109.685 60.8208 107.704 56.3361 105.062 52.3848C105.379 52.0948 105.979 51.5612 106.149 51.4043C106.199 50.8517 106.156 50.2722 106.725 49.6603C107.931 48.523 109.451 47.5799 111.284 46.4423C112.154 45.9279 112.959 45.5964 113.832 44.9484C114.03 44.8019 114.299 44.5699 114.507 44.4022C115.977 43.2237 116.316 41.1911 115.26 39.8614C114.204 38.5317 112.159 38.4065 110.688 39.585C110.479 39.7516 110.195 39.9688 110.007 40.1312C109.184 40.8426 108.676 41.5452 107.983 42.2832C106.471 43.8259 105.222 45.113 103.85 46.0409C103.255 46.3885 102.385 46.2682 101.99 46.2449L100.824 47.0806C94.1753 40.0763 85.1235 35.5982 75.3766 34.7283C75.3494 34.3179 75.3137 33.5761 75.3046 33.3529C74.9056 32.9693 74.4235 32.6418 74.3024 31.813C74.1691 30.1569 74.3917 28.3752 74.6496 26.2258C74.7919 25.2216 75.0284 24.3873 75.0688 23.2973C75.078 23.0495 75.0632 22.69 75.0622 22.4221C75.062 20.5322 73.6898 18.9998 71.9968 19ZM68.1585 42.8886L67.248 59.0447L67.1825 59.0776C67.1214 60.5229 65.9375 61.677 64.4839 61.677C63.8884 61.677 63.3388 61.4849 62.8922 61.1571L62.866 61.1703L49.6807 51.7794C53.7331 47.7759 58.9164 44.8172 64.89 43.4546C65.9812 43.2056 67.0719 43.0209 68.1585 42.8886ZM75.8417 42.8886C82.8159 43.7504 89.2657 46.9232 94.2081 51.786L81.108 61.1176L81.0621 61.0979C79.8994 61.9512 78.2611 61.7394 77.3548 60.5978C76.9835 60.1301 76.7887 59.5801 76.7653 59.0249L76.7522 59.0184L75.8417 42.8886ZM44.8991 57.814L56.9382 68.633L56.9251 68.6988C58.0117 69.6479 58.172 71.2949 57.2657 72.4368C56.8944 72.9045 56.3975 73.2182 55.8639 73.3647L55.8508 73.4173L40.4188 77.8923C39.6334 70.6765 41.3261 63.6621 44.8991 57.814ZM99.0094 57.8206C100.798 60.7336 102.153 63.9871 102.959 67.5143C103.756 70.9991 103.956 74.4778 103.627 77.8397L88.1166 73.3515L88.1035 73.2857C86.7145 72.9043 85.8609 71.4848 86.1843 70.0611C86.3168 69.4778 86.6249 68.9844 87.0423 68.6198L87.0358 68.5869L99.0094 57.8206ZM69.5274 69.4688H74.4596L77.5251 73.3186L76.4247 78.1226L71.9968 80.2614L67.5558 78.1161L66.4554 73.312L69.5274 69.4688ZM85.3393 82.6437C85.5489 82.6331 85.7576 82.652 85.9616 82.6898L85.9878 82.6569L101.95 85.3682C99.6142 91.9624 95.144 97.675 89.1711 101.498L82.9747 86.4606L82.9944 86.4343C82.4252 85.1055 82.9948 83.5472 84.3044 82.9135C84.6397 82.7513 84.99 82.6614 85.3393 82.6437ZM58.5298 82.7095C59.748 82.7267 60.8406 83.5761 61.1237 84.822C61.2562 85.4052 61.1917 85.9831 60.973 86.4935L61.0189 86.5527L54.888 101.439C49.1559 97.7432 44.5904 92.2099 42.1481 85.4208L57.9731 82.7227L57.9993 82.7556C58.1763 82.7229 58.3558 82.7071 58.5298 82.7095ZM71.8986 89.2312C72.3229 89.2155 72.7534 89.303 73.1627 89.501C73.6992 89.7606 74.1136 90.1692 74.3745 90.6592H74.4334L82.2346 104.821C81.2221 105.162 80.1813 105.454 79.1167 105.697C73.1505 107.058 67.2032 106.645 61.818 104.802L69.5995 90.6658H69.6126C70.0795 89.7888 70.965 89.2656 71.8986 89.2312Z" fill="white" stroke="white" stroke-width="0.25"/>
<defs>
<linearGradient id="paint0_linear_681_2845" x1="10" y1="15.5" x2="144" y2="131.5" gradientUnits="userSpaceOnUse">
<stop stop-color="#4D87FF"/>
<stop offset="1" stop-color="#0547D0"/>
</linearGradient>
</defs>
</svg>

-    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "storageClass"], ["spec", "nodeGroups"], ["spec", "nodeGroups", "md0"], ["spec", "nodeGroups", "md0", "minReplicas"], ["spec", "nodeGroups", "md0", "maxReplicas"], ["spec", "nodeGroups", "md0", "instanceType"], ["spec", "nodeGroups", "md0", "ephemeralStorage"], ["spec", "nodeGroups", "md0", "roles"], ["spec", "nodeGroups", "md0", "resources"], ["spec", "nodeGroups", "md0", "gpus"], ["spec", "version"], ["spec", "host"], ["spec", "addons"], ["spec", "addons", "certManager"], ["spec", "addons", "certManager", "enabled"], ["spec", "addons", "certManager", "valuesOverride"], ["spec", "addons", "cilium"], ["spec", "addons", "cilium", "valuesOverride"], ["spec", "addons", "gatewayAPI"], ["spec", "addons", "gatewayAPI", "enabled"], ["spec", "addons", "ingressNginx"], ["spec", "addons", "ingressNginx", "enabled"], ["spec", "addons", "ingressNginx", "exposeMethod"], ["spec", "addons", "ingressNginx", "hosts"], ["spec", "addons", "ingressNginx", "valuesOverride"], ["spec", "addons", "gpuOperator"], ["spec", "addons", "gpuOperator", "enabled"], ["spec", "addons", "gpuOperator", "valuesOverride"], ["spec", "addons", "fluxcd"], ["spec", "addons", "fluxcd", "enabled"], ["spec", "addons", "fluxcd", "valuesOverride"], ["spec", "addons", "monitoringAgents"], ["spec", "addons", "monitoringAgents", "enabled"], ["spec", "addons", "monitoringAgents", "valuesOverride"], ["spec", "addons", "verticalPodAutoscaler"], ["spec", "addons", "verticalPodAutoscaler", "valuesOverride"], ["spec", "addons", "velero"], ["spec", "addons", "velero", "enabled"], ["spec", "addons", "velero", "valuesOverride"], ["spec", "addons", "coredns"], ["spec", "addons", "coredns", "valuesOverride"], ["spec", "controlPlane"], ["spec", "controlPlane", "replicas"], ["spec", "controlPlane", "apiServer"], ["spec", "controlPlane", "apiServer", "resources"], ["spec", "controlPlane", "apiServer", "resourcesPreset"], ["spec", "controlPlane", "controllerManager"], ["spec", "controlPlane", "controllerManager", "resources"], ["spec", "controlPlane", "controllerManager", "resourcesPreset"], ["spec", "controlPlane", "scheduler"], ["spec", "controlPlane", "scheduler", "resources"], ["spec", "controlPlane", "scheduler", "resourcesPreset"], ["spec", "controlPlane", "konnectivity"], ["spec", "controlPlane", "konnectivity", "server"], ["spec", "controlPlane", "konnectivity", "server", "resources"], ["spec", "controlPlane", "konnectivity", "server", "resourcesPreset"]]
+    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "storageClass"], ["spec", "nodeGroups"], ["spec", "nodeGroups", "md0"], ["spec", "nodeGroups", "md0", "minReplicas"], ["spec", "nodeGroups", "md0", "maxReplicas"], ["spec", "nodeGroups", "md0", "instanceType"], ["spec", "nodeGroups", "md0", "ephemeralStorage"], ["spec", "nodeGroups", "md0", "roles"], ["spec", "nodeGroups", "md0", "resources"], ["spec", "nodeGroups", "md0", "gpus"], ["spec", "version"], ["spec", "host"], ["spec", "addons"], ["spec", "addons", "certManager"], ["spec", "addons", "certManager", "enabled"], ["spec", "addons", "certManager", "valuesOverride"], ["spec", "addons", "cilium"], ["spec", "addons", "cilium", "valuesOverride"], ["spec", "addons", "gatewayAPI"], ["spec", "addons", "gatewayAPI", "enabled"], ["spec", "addons", "ingressNginx"], ["spec", "addons", "ingressNginx", "enabled"], ["spec", "addons", "ingressNginx", "exposeMethod"], ["spec", "addons", "ingressNginx", "hosts"], ["spec", "addons", "ingressNginx", "valuesOverride"], ["spec", "addons", "gpuOperator"], ["spec", "addons", "gpuOperator", "enabled"], ["spec", "addons", "gpuOperator", "valuesOverride"], ["spec", "addons", "fluxcd"], ["spec", "addons", "fluxcd", "enabled"], ["spec", "addons", "fluxcd", "valuesOverride"], ["spec", "addons", "monitoringAgents"], ["spec", "addons", "monitoringAgents", "enabled"], ["spec", "addons", "monitoringAgents", "valuesOverride"], ["spec", "addons", "verticalPodAutoscaler"], ["spec", "addons", "verticalPodAutoscaler", "valuesOverride"], ["spec", "addons", "velero"], ["spec", "addons", "velero", "enabled"], ["spec", "addons", "velero", "valuesOverride"], ["spec", "addons", "coredns"], ["spec", "addons", "coredns", "valuesOverride"], ["spec", "controlPlane"], ["spec", "controlPlane", "replicas"], ["spec", "controlPlane", "apiServer"], ["spec", "controlPlane", "apiServer", "resources"], ["spec", "controlPlane", "apiServer", "resourcesPreset"], ["spec", "controlPlane", "controllerManager"], ["spec", "controlPlane", "controllerManager", "resources"], ["spec", "controlPlane", "controllerManager", "resourcesPreset"], ["spec", "controlPlane", "scheduler"], ["spec", "controlPlane", "scheduler", "resources"], ["spec", "controlPlane", "scheduler", "resourcesPreset"], ["spec", "controlPlane", "konnectivity"], ["spec", "controlPlane", "konnectivity", "server"], ["spec", "controlPlane", "konnectivity", "server", "resources"], ["spec", "controlPlane", "konnectivity", "server", "resourcesPreset"], ["spec", "images"], ["spec", "images", "waitForKubeconfig"]]
   secrets:
     exclude: []
     include:

From fb1ef59287114fe9ef6eec2351875139d4d84e97 Mon Sep 17 00:00:00 2001
From: Aleksei Sviridkin <f@lex.la>
Date: Mon, 27 Apr 2026 13:50:18 +0300
Subject: [PATCH 74/82] fix(kubernetes): drop undocumented status-beacon
 annotation from awaiting-etcd ConfigMap

The cozystack.io/status-beacon: "true" annotation had no consumer in
the chart, no documented contract, and no convention defined for other
charts to follow. It would have become accidental precedent for
contributors copying the pattern without understanding it.

The ConfigMap itself is self-explanatory: the name <release>-awaiting-etcd,
data.status: "awaiting-etcd", and the human-readable message in
data.message all surface the same operator signal via kubectl get cm.
Drop the annotation; keep the ConfigMap.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Aleksei Sviridkin <f@lex.la>
---
 packages/apps/kubernetes/templates/cluster.yaml | 2 --
 1 file changed, 2 deletions(-)

diff --git a/packages/apps/kubernetes/templates/cluster.yaml b/packages/apps/kubernetes/templates/cluster.yaml
index 12da9818..e1494e64 100644
--- a/packages/apps/kubernetes/templates/cluster.yaml
+++ b/packages/apps/kubernetes/templates/cluster.yaml
@@ -106,8 +106,6 @@ metadata:
     app.kubernetes.io/managed-by: {{ .Release.Service }}
     app.kubernetes.io/name: kubernetes
     app.kubernetes.io/instance: {{ .Release.Name }}
-  annotations:
-    cozystack.io/status-beacon: "true"
 data:
   status: "awaiting-etcd"
   message: |

From 7443e223453032d0a907311f3b5865889b713299 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Mon, 27 Apr 2026 17:18:45 +0500
Subject: [PATCH 75/82] build(linstor): include linstor-gui in root image build
 target

The linstor-gui package was added in #2382 with its own per-package
Makefile and Dockerfile, but the root Makefile's `build:` target was
not updated to invoke it. As a result `ghcr.io/cozystack/cozystack/
linstor-gui` has never been published (registry returns NAME_UNKNOWN)
and the chart's `image.tag` was never digest-pinned. Any cluster
deploying the chart hits ImagePullBackOff.

Wire the package into the root build alongside the other system
images. The next CI build will publish the image and the per-package
Makefile will rewrite values.yaml `image.repository`/`image.tag` to a
digest-pinned reference automatically.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 Makefile | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Makefile b/Makefile
index 0e94c5e6..5041ac64 100644
--- a/Makefile
+++ b/Makefile
@@ -22,6 +22,7 @@ build: build-deps
 	make -C packages/system/lineage-controller-webhook image
 	make -C packages/system/cilium image
 	make -C packages/system/linstor image
+	make -C packages/system/linstor-gui image
 	make -C packages/system/kubeovn-webhook image
 	make -C packages/system/kubeovn-plunger image
 	make -C packages/system/dashboard image

From a9a66bf066360c340203508fd6072977ff4eb8ea Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Mon, 27 Apr 2026 17:31:07 +0500
Subject: [PATCH 76/82] build(linstor): use shared BUILDX_ARGS for linstor-gui
 image build

The per-package Makefile added in #2382 hardcoded buildx flags
(--provenance, --builder, --platform=linux/amd64,linux/arm64,
--push, --load, --label) instead of using the shared $(BUILDX_ARGS)
macro from hack/common-envs.mk.

This broke CI: the runner's default docker driver does not support
multi-platform builds, and the hardcoded multi-arch platform list
crashed `make build` with "Multi-platform build is not supported for
the docker driver."

Replace the hardcoded flags with $(BUILDX_ARGS) to match every other
package (e.g. linstor, dashboard, cilium). $(BUILDX_ARGS) injects
--push, --load, --label, --provenance=false, and only sets --builder
or --platform when the operator explicitly exports BUILDER/PLATFORM.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 packages/system/linstor-gui/Makefile | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/packages/system/linstor-gui/Makefile b/packages/system/linstor-gui/Makefile
index e202f4cc..b408491d 100644
--- a/packages/system/linstor-gui/Makefile
+++ b/packages/system/linstor-gui/Makefile
@@ -12,18 +12,13 @@ image: image-linstor-gui
 
 image-linstor-gui:
 	docker buildx build images/linstor-gui \
-		--provenance false \
-		--builder=$(BUILDER) \
-		--platform=linux/amd64,linux/arm64 \
 		--build-arg LINSTOR_GUI_VERSION=$(LINSTOR_GUI_VERSION) \
 		--tag $(REGISTRY)/linstor-gui:$(call settag,$(LINSTOR_GUI_VERSION)) \
 		--tag $(REGISTRY)/linstor-gui:$(call settag,$(LINSTOR_GUI_VERSION)-$(TAG)) \
 		--cache-from type=registry,ref=$(REGISTRY)/linstor-gui:latest \
 		--cache-to type=inline \
 		--metadata-file images/linstor-gui.json \
-		--push=$(PUSH) \
-		--label "org.opencontainers.image.source=https://github.com/cozystack/cozystack" \
-		--load=$(LOAD)
+		$(BUILDX_ARGS)
 	REPOSITORY="$(REGISTRY)/linstor-gui" \
 		yq -i '.image.repository = strenv(REPOSITORY)' values.yaml
 	TAG="$(call settag,$(LINSTOR_GUI_VERSION))@$$(yq e '."containerimage.digest"' images/linstor-gui.json -o json -r)" \

From 073fb1630d67ace6655deb262d1734fab98e8ce4 Mon Sep 17 00:00:00 2001
From: Andrei Kvapil <kvapss@gmail.com>
Date: Mon, 27 Apr 2026 11:32:17 +0200
Subject: [PATCH 77/82] feat(linstor): bump linstor-csi to v1.10.6 with
 Protocol-C dual-attach fix

Bump linstor-csi to v1.10.6 and add an out-of-tree patch that overrides
DrbdOptions/Net/protocol to C on the resource-definition whenever a
volume is attached to a second node with allow-two-primaries (the
live-migration flow), and reverts it on detach.

Without the patch, any KubeVirt live migration of a VM whose volume
sits on a Protocol-A/B resource (for example a "replicated-async"
StorageClass) ends up in a permanent evacuation loop: drbdadm adjust
on the satellite rejects allow-two-primaries with "Protocol C
required" (errno 139), the migration fails, and KubeVirt retries
until manual intervention.

The override is set at the resource-definition level so it covers
every connection, including diskless TieBreaker peers, and is tagged
with an Aux marker so Detach reverts only the override we installed.
The 001 relocate-after-clone patch was regenerated against v1.10.6
(context shift only, no logic change).

Verified end-to-end on dev5: live migration of a KubeVirt VM whose
PVC sits on a Protocol-A resource-group now succeeds in a single
Attach.

Patch is upstreamed as draft PR piraeusdatastore/linstor-csi#435.

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Andrei Kvapil <kvapss@gmail.com>
---
 packages/system/linstor/Makefile              |   2 +-
 .../linstor/images/linstor-csi/Dockerfile     |   2 +-
 .../001-relocate-after-clone-restore.diff     |  62 +-------
 ...2-protocol-c-override-for-dual-attach.diff | 132 ++++++++++++++++++
 4 files changed, 140 insertions(+), 58 deletions(-)
 create mode 100644 packages/system/linstor/images/linstor-csi/patches/002-protocol-c-override-for-dual-attach.diff

diff --git a/packages/system/linstor/Makefile b/packages/system/linstor/Makefile
index 2ad4def9..8d566678 100644
--- a/packages/system/linstor/Makefile
+++ b/packages/system/linstor/Makefile
@@ -5,7 +5,7 @@ include ../../../hack/common-envs.mk
 include ../../../hack/package.mk
 
 LINSTOR_VERSION ?= 1.33.2
-LINSTOR_CSI_VERSION ?= v1.10.5
+LINSTOR_CSI_VERSION ?= v1.10.6
 
 image: image-piraeus-server image-linstor-csi
 
diff --git a/packages/system/linstor/images/linstor-csi/Dockerfile b/packages/system/linstor/images/linstor-csi/Dockerfile
index 6cfd44aa..df5d179d 100644
--- a/packages/system/linstor/images/linstor-csi/Dockerfile
+++ b/packages/system/linstor/images/linstor-csi/Dockerfile
@@ -1,6 +1,6 @@
 FROM golang:1.25 AS builder
 
-ARG VERSION=v1.10.5
+ARG VERSION=v1.10.6
 ARG LINSTOR_WAIT_UNTIL_VERSION=v0.3.1
 ARG TARGETARCH
 ARG TARGETOS
diff --git a/packages/system/linstor/images/linstor-csi/patches/001-relocate-after-clone-restore.diff b/packages/system/linstor/images/linstor-csi/patches/001-relocate-after-clone-restore.diff
index ff0aec0c..0e272670 100644
--- a/packages/system/linstor/images/linstor-csi/patches/001-relocate-after-clone-restore.diff
+++ b/packages/system/linstor/images/linstor-csi/patches/001-relocate-after-clone-restore.diff
@@ -1,25 +1,7 @@
 diff --git a/pkg/client/linstor.go b/pkg/client/linstor.go
-index f544493..98e7fde 100644
+index ac4651d..46d565d 100644
 --- a/pkg/client/linstor.go
 +++ b/pkg/client/linstor.go
-@@ -181,7 +181,7 @@ func LogLevel(s string) func(*Linstor) error {
- func (s *Linstor) ListAllWithStatus(ctx context.Context) ([]volume.VolumeStatus, error) {
- 	var vols []volume.VolumeStatus
- 
--	resourcesByName := make(map[string][]lapi.Resource)
-+	resourcesByName := make(map[string][]lapi.ResourceWithVolumes)
- 
- 	resDefs, err := s.client.ResourceDefinitions.GetAll(ctx, lapi.RDGetAllRequest{WithVolumeDefinitions: true})
- 	if err != nil {
-@@ -194,7 +194,7 @@ func (s *Linstor) ListAllWithStatus(ctx context.Context) ([]volume.VolumeStatus,
- 	}
- 
- 	for i := range allResources {
--		resourcesByName[allResources[i].Name] = append(resourcesByName[allResources[i].Name], allResources[i].Resource)
-+		resourcesByName[allResources[i].Name] = append(resourcesByName[allResources[i].Name], allResources[i])
- 	}
- 
- 	for _, rd := range resDefs {
 @@ -462,6 +462,14 @@ func (s *Linstor) Clone(ctx context.Context, vol, src *volume.Info, params *volu
  		return err
  	}
@@ -35,7 +17,7 @@ index f544493..98e7fde 100644
  	logger.Debug("reconcile extra properties")
  
  	err = s.client.ResourceDefinitions.Modify(ctx, vol.ID, lapi.GenericPropsModify{OverrideProps: vol.Properties})
-@@ -1280,6 +1288,14 @@ func (s *Linstor) VolFromSnap(ctx context.Context, snap *volume.Snapshot, vol *v
+@@ -1297,6 +1305,14 @@ func (s *Linstor) VolFromSnap(ctx context.Context, snap *volume.Snapshot, vol *v
  		return err
  	}
  
@@ -50,7 +32,7 @@ index f544493..98e7fde 100644
  	logger.Debug("reconcile extra properties")
  
  	err = s.client.ResourceDefinitions.Modify(ctx, vol.ID, lapi.GenericPropsModify{OverrideProps: vol.Properties})
-@@ -1470,9 +1486,8 @@ func (s *Linstor) reconcileSnapshotResources(ctx context.Context, snapshot *volu
+@@ -1487,9 +1503,8 @@ func (s *Linstor) reconcileSnapshotResources(ctx context.Context, snapshot *volu
  		return fmt.Errorf("snapshot '%s' not deployed on any node", snap.Name)
  	}
  
@@ -62,7 +44,7 @@ index f544493..98e7fde 100644
  	for _, snapNode := range snap.Nodes {
  		if err := s.NodeAvailable(ctx, snapNode); err != nil {
  			logger.WithField("selected node candidate", snapNode).WithError(err).Debug("node is not available")
-@@ -1480,13 +1495,23 @@ func (s *Linstor) reconcileSnapshotResources(ctx context.Context, snapshot *volu
+@@ -1497,13 +1512,23 @@ func (s *Linstor) reconcileSnapshotResources(ctx context.Context, snapshot *volu
  		}
  
  		if slices.Contains(preferredNodes, snapNode) {
@@ -92,7 +74,7 @@ index f544493..98e7fde 100644
  	}
  
  	if selectedNode == "" {
-@@ -1679,6 +1704,114 @@ func (s *Linstor) reconcileResourcePlacement(ctx context.Context, vol *volume.In
+@@ -1696,6 +1721,114 @@ func (s *Linstor) reconcileResourcePlacement(ctx context.Context, vol *volume.In
  	return nil
  }
  
@@ -207,40 +189,8 @@ index f544493..98e7fde 100644
  // FindSnapsByID searches the snapshot in the backend
  func (s *Linstor) FindSnapsByID(ctx context.Context, id string) ([]*volume.Snapshot, error) {
  	snapshotId, err := volume.ParseSnapshotId(id)
-@@ -2173,7 +2306,7 @@ func (s *Linstor) Status(ctx context.Context, volId string) ([]string, *csi.Volu
- 		"volume": volId,
- 	}).Debug("getting assignments")
- 
--	ress, err := s.client.Resources.GetAll(ctx, volId)
-+	ress, err := s.client.Resources.GetResourceView(ctx, &lapi.ListOpts{Resource: []string{volId}})
- 	if err != nil {
- 		return nil, nil, fmt.Errorf("failed to list resources for '%s': %w", volId, err)
- 	}
-@@ -2261,13 +2394,20 @@ func GetSnapshotRemoteAndReadiness(snap *lapi.Snapshot) (string, bool, error) {
- 	return "", slices.Contains(snap.Flags, lapiconsts.FlagSuccessful), nil
- }
- 
--func NodesAndConditionFromResources(ress []lapi.Resource) ([]string, *csi.VolumeCondition) {
-+func NodesAndConditionFromResources(ress []lapi.ResourceWithVolumes) ([]string, *csi.VolumeCondition) {
- 	var allNodes, abnormalNodes []string
- 
- 	for i := range ress {
- 		res := &ress[i]
- 
--		allNodes = append(allNodes, res.NodeName)
-+		// A resource is a CSI publish target if any of its volumes were created
-+		// by ControllerPublishVolume, identified by the temporary-diskless-attach property.
-+		if slices.ContainsFunc(res.Volumes, func(v lapi.Volume) bool {
-+			createdFor, ok := v.Props[linstor.PropertyCreatedFor]
-+			return ok && createdFor == linstor.CreatedForTemporaryDisklessAttach
-+		}) {
-+			allNodes = append(allNodes, res.NodeName)
-+		}
- 
- 		if res.State == nil {
- 			abnormalNodes = append(abnormalNodes, res.NodeName)
 diff --git a/pkg/volume/parameter.go b/pkg/volume/parameter.go
-index 39acd95..aed18ab 100644
+index 39acd95..54d1dfb 100644
 --- a/pkg/volume/parameter.go
 +++ b/pkg/volume/parameter.go
 @@ -50,6 +50,7 @@ const (
diff --git a/packages/system/linstor/images/linstor-csi/patches/002-protocol-c-override-for-dual-attach.diff b/packages/system/linstor/images/linstor-csi/patches/002-protocol-c-override-for-dual-attach.diff
new file mode 100644
index 00000000..279c636b
--- /dev/null
+++ b/packages/system/linstor/images/linstor-csi/patches/002-protocol-c-override-for-dual-attach.diff
@@ -0,0 +1,132 @@
+diff --git a/pkg/client/linstor.go b/pkg/client/linstor.go
+index ac4651d..9d75c79 100644
+--- a/pkg/client/linstor.go
++++ b/pkg/client/linstor.go
+@@ -653,11 +653,35 @@ func (s *Linstor) Attach(ctx context.Context, volId, node string, rwxBlock bool)
+ 	}
+ 
+ 	if otherResInUse > 0 && rwxBlock {
+-		rdPropsModify := lapi.GenericPropsModify{OverrideProps: map[string]string{
++		rdProps := map[string]string{
+ 			linstor.PropertyAllowTwoPrimaries: "yes",
+-		}}
++		}
+ 
+-		err = s.client.ResourceDefinitions.Modify(ctx, volId, rdPropsModify)
++		// DRBD requires Protocol C whenever allow-two-primaries is enabled.
++		// If the resource is configured with Protocol A or B (typically
++		// through its resource-group), drbdadm adjust on the satellites
++		// fails with "Protocol C required" (errno 139) and live migration
++		// cannot proceed. Override Protocol to C on the resource-definition
++		// itself so it applies to every connection (including diskless
++		// peers, e.g. a TieBreaker). The marker lets Detach revert exactly
++		// the override we installed without disturbing operator-set props.
++		proto, protoErr := s.getEffectiveDrbdProtocol(ctx, volId)
++		if protoErr != nil {
++			s.log.WithError(protoErr).WithField("volume", volId).
++				Warn("failed to determine effective DRBD protocol; skipping Protocol=C override for dual-attach")
++		} else if proto == "A" || proto == "B" {
++			s.log.WithFields(logrus.Fields{
++				"volume":   volId,
++				"protocol": proto,
++			}).Info("installing Protocol=C override on resource-definition for dual-attach")
++
++			rdProps[linstor.PropertyDrbdNetProtocol] = "C"
++			rdProps[linstor.PropertyCsiProtocolOverride] = "yes"
++		}
++
++		err = s.client.ResourceDefinitions.Modify(ctx, volId, lapi.GenericPropsModify{
++			OverrideProps: rdProps,
++		})
+ 		if err != nil {
+ 			return "", err
+ 		}
+@@ -774,11 +798,26 @@ func (s *Linstor) Detach(ctx context.Context, volId, node string) error {
+ 	}
+ 
+ 	if resInUse == 1 {
+-		rdPropsModify := lapi.GenericPropsModify{DeleteProps: []string{
+-			linstor.PropertyAllowTwoPrimaries,
+-		}}
++		deleteProps := []string{linstor.PropertyAllowTwoPrimaries}
++
++		// Drop the Protocol=C override only if Attach installed it (gated by
++		// our marker), so an operator-set Protocol property on the
++		// resource-definition is preserved.
++		rd, getErr := s.client.ResourceDefinitions.Get(ctx, volId)
++		if getErr != nil {
++			log.WithError(getErr).Warn("failed to fetch resource-definition props; skipping Protocol=C override removal")
++		} else if rd.Props[linstor.PropertyCsiProtocolOverride] != "" {
++			log.Info("removing Protocol=C override from resource-definition")
+ 
+-		err = s.client.ResourceDefinitions.Modify(ctx, volId, rdPropsModify)
++			deleteProps = append(deleteProps,
++				linstor.PropertyDrbdNetProtocol,
++				linstor.PropertyCsiProtocolOverride,
++			)
++		}
++
++		err = s.client.ResourceDefinitions.Modify(ctx, volId, lapi.GenericPropsModify{
++			DeleteProps: deleteProps,
++		})
+ 		if err != nil {
+ 			return nil404(err)
+ 		}
+@@ -805,6 +844,33 @@ func (s *Linstor) Detach(ctx context.Context, volId, node string) error {
+ 	return nil404(s.client.Resources.Delete(ctx, volId, node))
+ }
+ 
++// getEffectiveDrbdProtocol returns the DRBD network protocol ("A", "B" or "C")
++// that LINSTOR will hand to drbdadm for the given resource. Resource-definition
++// properties take precedence; otherwise the value is inherited from the
++// resource-group. An empty string is returned when neither level sets the
++// property (LINSTOR's compiled-in default is "C").
++func (s *Linstor) getEffectiveDrbdProtocol(ctx context.Context, volId string) (string, error) {
++	rd, err := s.client.ResourceDefinitions.Get(ctx, volId)
++	if err != nil {
++		return "", err
++	}
++
++	if v, ok := rd.Props[linstor.PropertyDrbdNetProtocol]; ok {
++		return v, nil
++	}
++
++	if rd.ResourceGroupName == "" {
++		return "", nil
++	}
++
++	rg, err := s.client.ResourceGroups.Get(ctx, rd.ResourceGroupName)
++	if err != nil {
++		return "", err
++	}
++
++	return rg.Props[linstor.PropertyDrbdNetProtocol], nil
++}
++
+ // CapacityBytes returns the amount of free space in the storage pool specified by the params and topology.
+ func (s *Linstor) CapacityBytes(ctx context.Context, storagePools []string, overProvision *float64, segments map[string]string) (int64, error) {
+ 	log := s.log.WithField("storage-pools", storagePools).WithField("segments", segments)
+diff --git a/pkg/linstor/const.go b/pkg/linstor/const.go
+index 9a5f79c..c8bc9c3 100644
+--- a/pkg/linstor/const.go
++++ b/pkg/linstor/const.go
+@@ -44,6 +44,19 @@ const (
+ 	// PropertyAllowTwoPrimaries is DRBD option to allow second primary. Mainly used for live-migration.
+ 	PropertyAllowTwoPrimaries = lc.NamespcDrbdNetOptions + "/allow-two-primaries"
+ 
++	// PropertyDrbdNetProtocol is the DRBD network replication protocol option
++	// (A = async, B = semi-sync, C = sync). DRBD requires Protocol C whenever
++	// allow-two-primaries is enabled, otherwise drbdadm adjust fails with
++	// "Protocol C required" (errno 139).
++	PropertyDrbdNetProtocol = lc.NamespcDrbdNetOptions + "/protocol"
++
++	// PropertyCsiProtocolOverride marks a resource-connection where this driver
++	// has installed a temporary Protocol=C override to make a Protocol-A/B
++	// resource compatible with allow-two-primaries during live migration. The
++	// marker lets us safely remove only the overrides we set, without touching
++	// connection-level overrides installed by the operator.
++	PropertyCsiProtocolOverride = lc.NamespcAuxiliary + "/csi-protocol-override"
++
+ 	// CreatedForTemporaryDisklessAttach marks a resource as temporary, i.e. it should be removed after it is no longer
+ 	// needed.
+ 	CreatedForTemporaryDisklessAttach = "temporary-diskless-attach"

From 80631bc9160a4567cce8facdf781799f15455497 Mon Sep 17 00:00:00 2001
From: mattia-eleuteri <mattia@hidora.io>
Date: Mon, 27 Apr 2026 16:29:05 +0200
Subject: [PATCH 78/82] [vm-instance] Set wholeIP annotation conditionally on
 externalMethod
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Render `networking.cozystack.io/wholeIP: "false"` on the Service when
`externalMethod: PortList` is configured (was always `"true"` before).
Combined with cozy-proxy v0.3.0+ which adds per-port filtering for
"false"-annotated services, this makes `externalMethod: PortList`
behave as documented: only ports listed in `externalPorts` are
reachable from the LoadBalancer IP.

Backward-compatible: existing services with `externalMethod: WholeIP`
continue to set `wholeIP: "true"` and behave identically. cozy-proxy
versions older than v0.3.0 ignore Services with `wholeIP: "false"`,
which means PortList Services on older cozy-proxy will lose their
egress IP preservation — but that path was already non-functional
ingress-wise, so this is not a regression for users actually relying
on PortList.

Signed-off-by: mattia-eleuteri <mattia@hidora.io>
---
 docs/changelogs/v1.3.1.md                        | 11 +++++++++++
 packages/apps/vm-instance/templates/service.yaml |  2 +-
 2 files changed, 12 insertions(+), 1 deletion(-)
 create mode 100644 docs/changelogs/v1.3.1.md

diff --git a/docs/changelogs/v1.3.1.md b/docs/changelogs/v1.3.1.md
new file mode 100644
index 00000000..7f9ecfab
--- /dev/null
+++ b/docs/changelogs/v1.3.1.md
@@ -0,0 +1,11 @@
+<!--
+https://github.com/cozystack/cozystack/releases/tag/v1.3.1
+-->
+
+## Bug Fixes
+
+* **[vm-instance] Fix `externalMethod: PortList` not filtering ingress ports**: The `vm-instance` chart now sets `networking.cozystack.io/wholeIP: "false"` on the rendered Service when `externalMethod: PortList` is configured, signaling cozy-proxy to install per-port ingress filtering based on `Service.spec.ports`. Previously the annotation was always `"true"`, which kept cozy-proxy in whole-IP passthrough mode and made `PortList` non-functional. **Requires cozy-proxy v0.3.0 or later** (bundled with this chart). ([@mattia-eleuteri](https://github.com/mattia-eleuteri) in #<TBD>).
+
+---
+
+**Full Changelog**: https://github.com/cozystack/cozystack/compare/v1.3.0...v1.3.1
diff --git a/packages/apps/vm-instance/templates/service.yaml b/packages/apps/vm-instance/templates/service.yaml
index b7e3a420..4607ffcf 100644
--- a/packages/apps/vm-instance/templates/service.yaml
+++ b/packages/apps/vm-instance/templates/service.yaml
@@ -9,7 +9,7 @@ metadata:
 {{- if .Values.external }}
     service.kubernetes.io/service-proxy-name: "cozy-proxy"
   annotations:
-    networking.cozystack.io/wholeIP: "true"
+    networking.cozystack.io/wholeIP: {{ ternary "true" "false" (eq .Values.externalMethod "WholeIP") | quote }}
 {{- end }}
 spec:
   type: {{ ternary "LoadBalancer" "ClusterIP" .Values.external }}

From b0afc9a07c8a842915af1fa98ff1e7d17ccb39c8 Mon Sep 17 00:00:00 2001
From: mattia-eleuteri <mattia@hidora.io>
Date: Tue, 28 Apr 2026 08:37:13 +0200
Subject: [PATCH 79/82] [vm-instance] Add externalAllowICMP knob, drop in-PR
 changelog

- Add `externalAllowICMP` value (default true) propagated as
  `networking.cozystack.io/allowICMP` annotation on the rendered Service
  when `externalMethod: PortList`. The cozy-proxy companion (released as
  part of cozystack/cozy-proxy#11 + #12) drops ICMP by default in
  port-filter mode, which breaks ping and PMTU discovery; defaulting the
  chart to "true" preserves user expectations while still allowing
  operators to opt out by setting `externalAllowICMP: false`.

- Remove the v1.3.1.md changelog entry. Project convention is to add
  changelogs in a dedicated "docs: add changelog for vX.Y.Z" commit at
  release time, not as part of feature/fix PRs.

Signed-off-by: mattia-eleuteri <mattia@hidora.io>
---
 api/apps/v1alpha1/vminstance/types.go         |  3 ++
 docs/changelogs/v1.3.1.md                     | 11 ----
 packages/apps/vm-instance/README.md           | 51 ++++++++++---------
 .../apps/vm-instance/templates/service.yaml   |  3 ++
 packages/apps/vm-instance/values.schema.json  |  5 ++
 packages/apps/vm-instance/values.yaml         |  3 ++
 .../vm-instance-rd/cozyrds/vm-instance.yaml   |  4 +-
 7 files changed, 42 insertions(+), 38 deletions(-)
 delete mode 100644 docs/changelogs/v1.3.1.md

diff --git a/api/apps/v1alpha1/vminstance/types.go b/api/apps/v1alpha1/vminstance/types.go
index 2bb059c6..9c48724a 100644
--- a/api/apps/v1alpha1/vminstance/types.go
+++ b/api/apps/v1alpha1/vminstance/types.go
@@ -26,6 +26,9 @@ type ConfigSpec struct {
 	// Ports to forward from outside the cluster.
 	// +kubebuilder:default:={22}
 	ExternalPorts []int `json:"externalPorts,omitempty"`
+	// Whether to accept ICMP traffic to the VM in PortList mode (preserves ping and PMTU discovery). No effect in WholeIP mode. Default true so ping behaves as users expect even when port filtering is in effect.
+	// +kubebuilder:default:=true
+	ExternalAllowICMP bool `json:"externalAllowICMP"`
 	// Requested running state of the VirtualMachineInstance
 	// +kubebuilder:default:="Always"
 	RunStrategy RunStrategy `json:"runStrategy"`
diff --git a/docs/changelogs/v1.3.1.md b/docs/changelogs/v1.3.1.md
deleted file mode 100644
index 7f9ecfab..00000000
--- a/docs/changelogs/v1.3.1.md
+++ /dev/null
@@ -1,11 +0,0 @@
-<!--
-https://github.com/cozystack/cozystack/releases/tag/v1.3.1
--->
-
-## Bug Fixes
-
-* **[vm-instance] Fix `externalMethod: PortList` not filtering ingress ports**: The `vm-instance` chart now sets `networking.cozystack.io/wholeIP: "false"` on the rendered Service when `externalMethod: PortList` is configured, signaling cozy-proxy to install per-port ingress filtering based on `Service.spec.ports`. Previously the annotation was always `"true"`, which kept cozy-proxy in whole-IP passthrough mode and made `PortList` non-functional. **Requires cozy-proxy v0.3.0 or later** (bundled with this chart). ([@mattia-eleuteri](https://github.com/mattia-eleuteri) in #<TBD>).
-
----
-
-**Full Changelog**: https://github.com/cozystack/cozystack/compare/v1.3.0...v1.3.1
diff --git a/packages/apps/vm-instance/README.md b/packages/apps/vm-instance/README.md
index a2b6603e..9d6a52b2 100644
--- a/packages/apps/vm-instance/README.md
+++ b/packages/apps/vm-instance/README.md
@@ -36,31 +36,32 @@ virtctl ssh <user>@<vm>
 
 ### Common parameters
 
-| Name                | Description                                                                                                                       | Type       | Value       |
-| ------------------- | --------------------------------------------------------------------------------------------------------------------------------- | ---------- | ----------- |
-| `external`          | Enable external access from outside the cluster.                                                                                  | `bool`     | `false`     |
-| `externalMethod`    | Method to pass through traffic to the VM.                                                                                         | `string`   | `PortList`  |
-| `externalPorts`     | Ports to forward from outside the cluster.                                                                                        | `[]int`    | `[22]`      |
-| `runStrategy`       | Requested running state of the VirtualMachineInstance                                                                             | `string`   | `Always`    |
-| `instanceType`      | Virtual Machine instance type.                                                                                                    | `string`   | `u1.medium` |
-| `instanceProfile`   | Virtual Machine preferences profile.                                                                                              | `string`   | `ubuntu`    |
-| `disks`             | List of disks to attach.                                                                                                          | `[]object` | `[]`        |
-| `disks[i].name`     | Disk name.                                                                                                                        | `string`   | `""`        |
-| `disks[i].bus`      | Disk bus type (e.g. "sata").                                                                                                      | `string`   | `""`        |
-| `networks`          | Networks to attach the VM to.                                                                                                     | `[]object` | `[]`        |
-| `networks[i].name`  | Network attachment name.                                                                                                          | `string`   | `""`        |
-| `subnets`           | Deprecated: use networks instead.                                                                                                 | `[]object` | `[]`        |
-| `subnets[i].name`   | Network attachment name.                                                                                                          | `string`   | `""`        |
-| `gpus`              | List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).                                                               | `[]object` | `[]`        |
-| `gpus[i].name`      | The name of the GPU resource to attach.                                                                                           | `string`   | `""`        |
-| `cpuModel`          | Model specifies the CPU model inside the VMI. List of available models https://github.com/libvirt/libvirt/tree/master/src/cpu_map | `string`   | `""`        |
-| `resources`         | Resource configuration for the virtual machine.                                                                                   | `object`   | `{}`        |
-| `resources.cpu`     | Number of CPU cores allocated.                                                                                                    | `quantity` | `""`        |
-| `resources.memory`  | Amount of memory allocated.                                                                                                       | `quantity` | `""`        |
-| `resources.sockets` | Number of CPU sockets (vCPU topology).                                                                                            | `quantity` | `""`        |
-| `sshKeys`           | List of SSH public keys for authentication.                                                                                       | `[]string` | `[]`        |
-| `cloudInit`         | Cloud-init user data.                                                                                                             | `string`   | `""`        |
-| `cloudInitSeed`     | Seed string to generate SMBIOS UUID for the VM.                                                                                   | `string`   | `""`        |
+| Name                | Description                                                                                                                                                                                                   | Type       | Value       |
+| ------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------- | ----------- |
+| `external`          | Enable external access from outside the cluster.                                                                                                                                                              | `bool`     | `false`     |
+| `externalMethod`    | Method to pass through traffic to the VM.                                                                                                                                                                     | `string`   | `PortList`  |
+| `externalPorts`     | Ports to forward from outside the cluster.                                                                                                                                                                    | `[]int`    | `[22]`      |
+| `externalAllowICMP` | Whether to accept ICMP traffic to the VM in PortList mode (preserves ping and PMTU discovery). No effect in WholeIP mode. Default true so ping behaves as users expect even when port filtering is in effect. | `bool`     | `true`      |
+| `runStrategy`       | Requested running state of the VirtualMachineInstance                                                                                                                                                         | `string`   | `Always`    |
+| `instanceType`      | Virtual Machine instance type.                                                                                                                                                                                | `string`   | `u1.medium` |
+| `instanceProfile`   | Virtual Machine preferences profile.                                                                                                                                                                          | `string`   | `ubuntu`    |
+| `disks`             | List of disks to attach.                                                                                                                                                                                      | `[]object` | `[]`        |
+| `disks[i].name`     | Disk name.                                                                                                                                                                                                    | `string`   | `""`        |
+| `disks[i].bus`      | Disk bus type (e.g. "sata").                                                                                                                                                                                  | `string`   | `""`        |
+| `networks`          | Networks to attach the VM to.                                                                                                                                                                                 | `[]object` | `[]`        |
+| `networks[i].name`  | Network attachment name.                                                                                                                                                                                      | `string`   | `""`        |
+| `subnets`           | Deprecated: use networks instead.                                                                                                                                                                             | `[]object` | `[]`        |
+| `subnets[i].name`   | Network attachment name.                                                                                                                                                                                      | `string`   | `""`        |
+| `gpus`              | List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).                                                                                                                                           | `[]object` | `[]`        |
+| `gpus[i].name`      | The name of the GPU resource to attach.                                                                                                                                                                       | `string`   | `""`        |
+| `cpuModel`          | Model specifies the CPU model inside the VMI. List of available models https://github.com/libvirt/libvirt/tree/master/src/cpu_map                                                                             | `string`   | `""`        |
+| `resources`         | Resource configuration for the virtual machine.                                                                                                                                                               | `object`   | `{}`        |
+| `resources.cpu`     | Number of CPU cores allocated.                                                                                                                                                                                | `quantity` | `""`        |
+| `resources.memory`  | Amount of memory allocated.                                                                                                                                                                                   | `quantity` | `""`        |
+| `resources.sockets` | Number of CPU sockets (vCPU topology).                                                                                                                                                                        | `quantity` | `""`        |
+| `sshKeys`           | List of SSH public keys for authentication.                                                                                                                                                                   | `[]string` | `[]`        |
+| `cloudInit`         | Cloud-init user data.                                                                                                                                                                                         | `string`   | `""`        |
+| `cloudInitSeed`     | Seed string to generate SMBIOS UUID for the VM.                                                                                                                                                               | `string`   | `""`        |
 
 
 ## U Series
diff --git a/packages/apps/vm-instance/templates/service.yaml b/packages/apps/vm-instance/templates/service.yaml
index 4607ffcf..cfeffa81 100644
--- a/packages/apps/vm-instance/templates/service.yaml
+++ b/packages/apps/vm-instance/templates/service.yaml
@@ -10,6 +10,9 @@ metadata:
     service.kubernetes.io/service-proxy-name: "cozy-proxy"
   annotations:
     networking.cozystack.io/wholeIP: {{ ternary "true" "false" (eq .Values.externalMethod "WholeIP") | quote }}
+    {{- if eq .Values.externalMethod "PortList" }}
+    networking.cozystack.io/allowICMP: {{ ternary "true" "false" (ne .Values.externalAllowICMP false) | quote }}
+    {{- end }}
 {{- end }}
 spec:
   type: {{ ternary "LoadBalancer" "ClusterIP" .Values.external }}
diff --git a/packages/apps/vm-instance/values.schema.json b/packages/apps/vm-instance/values.schema.json
index 34f7f634..01bd30a9 100644
--- a/packages/apps/vm-instance/values.schema.json
+++ b/packages/apps/vm-instance/values.schema.json
@@ -26,6 +26,11 @@
         "type": "integer"
       }
     },
+    "externalAllowICMP": {
+      "description": "Whether to accept ICMP traffic to the VM in PortList mode (preserves ping and PMTU discovery). No effect in WholeIP mode. Default true so ping behaves as users expect even when port filtering is in effect.",
+      "type": "boolean",
+      "default": true
+    },
     "runStrategy": {
       "description": "Requested running state of the VirtualMachineInstance",
       "type": "string",
diff --git a/packages/apps/vm-instance/values.yaml b/packages/apps/vm-instance/values.yaml
index f07b75d3..92e399c2 100644
--- a/packages/apps/vm-instance/values.yaml
+++ b/packages/apps/vm-instance/values.yaml
@@ -31,6 +31,9 @@ externalMethod: PortList
 externalPorts:
 - 22
 
+## @param {bool} externalAllowICMP - Whether to accept ICMP traffic to the VM in PortList mode (preserves ping and PMTU discovery). No effect in WholeIP mode. Default true so ping behaves as users expect even when port filtering is in effect.
+externalAllowICMP: true
+
 ## @enum {string} RunStrategy - Requested running state of the VirtualMachineInstance
 ## @value Always - VMI should always be running
 ## @value Halted - VMI should never be running
diff --git a/packages/system/vm-instance-rd/cozyrds/vm-instance.yaml b/packages/system/vm-instance-rd/cozyrds/vm-instance.yaml
index e93fa86b..e56bd9dd 100644
--- a/packages/system/vm-instance-rd/cozyrds/vm-instance.yaml
+++ b/packages/system/vm-instance-rd/cozyrds/vm-instance.yaml
@@ -8,7 +8,7 @@ spec:
     singular: vminstance
     plural: vminstances
     openAPISchema: |-
-      {"title":"Chart Values","type":"object","properties":{"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"externalMethod":{"description":"Method to pass through traffic to the VM.","type":"string","default":"PortList","enum":["PortList","WholeIP"]},"externalPorts":{"description":"Ports to forward from outside the cluster.","type":"array","default":[22],"items":{"type":"integer"}},"runStrategy":{"description":"Requested running state of the VirtualMachineInstance","type":"string","default":"Always","enum":["Always","Halted","Manual","RerunOnFailure","Once"]},"instanceType":{"description":"Virtual Machine instance type.","type":"string","default":"u1.medium"},"instanceProfile":{"description":"Virtual Machine preferences profile.","type":"string","default":"ubuntu","enum":["alpine","centos.7","centos.7.desktop","centos.stream10","centos.stream10.desktop","centos.stream8","centos.stream8.desktop","centos.stream8.dpdk","centos.stream9","centos.stream9.desktop","centos.stream9.dpdk","cirros","fedora","fedora.arm64","opensuse.leap","opensuse.tumbleweed","rhel.10","rhel.10.arm64","rhel.7","rhel.7.desktop","rhel.8","rhel.8.desktop","rhel.8.dpdk","rhel.9","rhel.9.arm64","rhel.9.desktop","rhel.9.dpdk","rhel.9.realtime","sles","ubuntu","windows.10","windows.10.virtio","windows.11","windows.11.virtio","windows.2k16","windows.2k16.virtio","windows.2k19","windows.2k19.virtio","windows.2k22","windows.2k22.virtio","windows.2k25","windows.2k25.virtio",""]},"disks":{"description":"List of disks to attach.","type":"array","default":[],"items":{"type":"object","required":["name"],"properties":{"bus":{"description":"Disk bus type (e.g. \"sata\").","type":"string"},"name":{"description":"Disk name.","type":"string"}}}},"networks":{"description":"Networks to attach the VM to.","type":"array","default":[],"items":{"type":"object","properties":{"name":{"description":"Network attachment name.","type":"string"}}}},"subnets":{"description":"Deprecated: use networks instead.","type":"array","default":[],"items":{"type":"object","properties":{"name":{"description":"Network attachment name.","type":"string"}}}},"gpus":{"description":"List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).","type":"array","default":[],"items":{"type":"object","required":["name"],"properties":{"name":{"description":"The name of the GPU resource to attach.","type":"string"}}}},"cpuModel":{"description":"Model specifies the CPU model inside the VMI. List of available models https://github.com/libvirt/libvirt/tree/master/src/cpu_map","type":"string","default":""},"resources":{"description":"Resource configuration for the virtual machine.","type":"object","default":{},"properties":{"cpu":{"description":"Number of CPU cores allocated.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Amount of memory allocated.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"sockets":{"description":"Number of CPU sockets (vCPU topology).","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"sshKeys":{"description":"List of SSH public keys for authentication.","type":"array","default":[],"items":{"type":"string"}},"cloudInit":{"description":"Cloud-init user data.","type":"string","default":""},"cloudInitSeed":{"description":"Seed string to generate SMBIOS UUID for the VM.","type":"string","default":""}}}
+      {"title":"Chart Values","type":"object","properties":{"external":{"description":"Enable external access from outside the cluster.","type":"boolean","default":false},"externalMethod":{"description":"Method to pass through traffic to the VM.","type":"string","default":"PortList","enum":["PortList","WholeIP"]},"externalPorts":{"description":"Ports to forward from outside the cluster.","type":"array","default":[22],"items":{"type":"integer"}},"externalAllowICMP":{"description":"Whether to accept ICMP traffic to the VM in PortList mode (preserves ping and PMTU discovery). No effect in WholeIP mode. Default true so ping behaves as users expect even when port filtering is in effect.","type":"boolean","default":true},"runStrategy":{"description":"Requested running state of the VirtualMachineInstance","type":"string","default":"Always","enum":["Always","Halted","Manual","RerunOnFailure","Once"]},"instanceType":{"description":"Virtual Machine instance type.","type":"string","default":"u1.medium"},"instanceProfile":{"description":"Virtual Machine preferences profile.","type":"string","default":"ubuntu","enum":["alpine","centos.7","centos.7.desktop","centos.stream10","centos.stream10.desktop","centos.stream8","centos.stream8.desktop","centos.stream8.dpdk","centos.stream9","centos.stream9.desktop","centos.stream9.dpdk","cirros","fedora","fedora.arm64","opensuse.leap","opensuse.tumbleweed","rhel.10","rhel.10.arm64","rhel.7","rhel.7.desktop","rhel.8","rhel.8.desktop","rhel.8.dpdk","rhel.9","rhel.9.arm64","rhel.9.desktop","rhel.9.dpdk","rhel.9.realtime","sles","ubuntu","windows.10","windows.10.virtio","windows.11","windows.11.virtio","windows.2k16","windows.2k16.virtio","windows.2k19","windows.2k19.virtio","windows.2k22","windows.2k22.virtio","windows.2k25","windows.2k25.virtio",""]},"disks":{"description":"List of disks to attach.","type":"array","default":[],"items":{"type":"object","required":["name"],"properties":{"bus":{"description":"Disk bus type (e.g. \"sata\").","type":"string"},"name":{"description":"Disk name.","type":"string"}}}},"networks":{"description":"Networks to attach the VM to.","type":"array","default":[],"items":{"type":"object","properties":{"name":{"description":"Network attachment name.","type":"string"}}}},"subnets":{"description":"Deprecated: use networks instead.","type":"array","default":[],"items":{"type":"object","properties":{"name":{"description":"Network attachment name.","type":"string"}}}},"gpus":{"description":"List of GPUs to attach (NVIDIA driver requires at least 4 GiB RAM).","type":"array","default":[],"items":{"type":"object","required":["name"],"properties":{"name":{"description":"The name of the GPU resource to attach.","type":"string"}}}},"cpuModel":{"description":"Model specifies the CPU model inside the VMI. List of available models https://github.com/libvirt/libvirt/tree/master/src/cpu_map","type":"string","default":""},"resources":{"description":"Resource configuration for the virtual machine.","type":"object","default":{},"properties":{"cpu":{"description":"Number of CPU cores allocated.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"memory":{"description":"Amount of memory allocated.","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true},"sockets":{"description":"Number of CPU sockets (vCPU topology).","pattern":"^(\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\\+|-)?(([0-9]+(\\.[0-9]*)?)|(\\.[0-9]+))))?$","anyOf":[{"type":"integer"},{"type":"string"}],"x-kubernetes-int-or-string":true}}},"sshKeys":{"description":"List of SSH public keys for authentication.","type":"array","default":[],"items":{"type":"string"}},"cloudInit":{"description":"Cloud-init user data.","type":"string","default":""},"cloudInitSeed":{"description":"Seed string to generate SMBIOS UUID for the VM.","type":"string","default":""}}}
   release:
     prefix: vm-instance-
     labels:
@@ -26,7 +26,7 @@ spec:
     tags:
       - compute
     icon: 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
-    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "external"], ["spec", "externalMethod"], ["spec", "externalPorts"], ["spec", "runStrategy"], ["spec", "instanceType"], ["spec", "instanceProfile"], ["spec", "disks"], ["spec", "networks"], ["spec", "subnets"], ["spec", "gpus"], ["spec", "cpuModel"], ["spec", "resources"], ["spec", "sshKeys"], ["spec", "cloudInit"], ["spec", "cloudInitSeed"]]
+    keysOrder: [["apiVersion"], ["appVersion"], ["kind"], ["metadata"], ["metadata", "name"], ["spec", "external"], ["spec", "externalMethod"], ["spec", "externalPorts"], ["spec", "externalAllowICMP"], ["spec", "runStrategy"], ["spec", "instanceType"], ["spec", "instanceProfile"], ["spec", "disks"], ["spec", "networks"], ["spec", "subnets"], ["spec", "gpus"], ["spec", "cpuModel"], ["spec", "resources"], ["spec", "sshKeys"], ["spec", "cloudInit"], ["spec", "cloudInitSeed"]]
   secrets:
     exclude: []
     include: []

From d20285836cede23c3d7cfe091e5444359bac4320 Mon Sep 17 00:00:00 2001
From: Andrei Kvapil <kvapss@gmail.com>
Date: Tue, 28 Apr 2026 11:14:27 +0200
Subject: [PATCH 80/82] feat(cozy-proxy): bump to v0.3.0

Pulls in the per-port filtering and allowICMP support that the companion
vm-instance chart fix in #2501 relies on. cozy-proxy v0.3.0 also tightens
the selector to the standard service.kubernetes.io/service-proxy-name=cozy-proxy
label and switches the default ingress mode to port-filter; both are
already covered by the vm-instance chart (label landed in #2357,
wholeIP/allowICMP wired explicitly in #2501), so VM workloads upgrade
transparently.

Out-of-tree consumers using cozy-proxy annotations directly (without the
label, or relying on the absent-annotation passthrough default) are
called out in the upstream v0.3.0 release notes:
https://github.com/cozystack/cozy-proxy/releases/tag/v0.3.0

Signed-off-by: Andrei Kvapil <kvapss@gmail.com>
---
 packages/system/cozy-proxy/charts/cozy-proxy/Chart.yaml  | 4 ++--
 packages/system/cozy-proxy/charts/cozy-proxy/values.yaml | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/packages/system/cozy-proxy/charts/cozy-proxy/Chart.yaml b/packages/system/cozy-proxy/charts/cozy-proxy/Chart.yaml
index 13352680..58f166f6 100644
--- a/packages/system/cozy-proxy/charts/cozy-proxy/Chart.yaml
+++ b/packages/system/cozy-proxy/charts/cozy-proxy/Chart.yaml
@@ -2,5 +2,5 @@ apiVersion: v2
 name: cozy-proxy
 description: A simple kube-proxy addon for 1:1 NAT services in Kubernetes using an NFT backend
 type: application
-version: 0.2.0
-appVersion: 0.2.0
+version: 0.3.0
+appVersion: 0.3.0
diff --git a/packages/system/cozy-proxy/charts/cozy-proxy/values.yaml b/packages/system/cozy-proxy/charts/cozy-proxy/values.yaml
index 8cde5bed..e143e926 100644
--- a/packages/system/cozy-proxy/charts/cozy-proxy/values.yaml
+++ b/packages/system/cozy-proxy/charts/cozy-proxy/values.yaml
@@ -1,6 +1,6 @@
 image:
   repository: ghcr.io/cozystack/cozystack/cozy-proxy
-  tag: v0.2.0
+  tag: v0.3.0
   pullPolicy: IfNotPresent
 
 daemonset:

From 7257b6aed4bb2b6499168eab467a400196f2e625 Mon Sep 17 00:00:00 2001
From: IvanHunters <xorokhotnikov@gmail.com>
Date: Tue, 28 Apr 2026 12:26:57 +0300
Subject: [PATCH 81/82] fix(api): address review feedback on TenantNamespace
 RBAC

Address review feedback from sircthulhu and CodeRabbit:

1. Return Forbidden instead of NotFound for unauthorized access
   - Get() now returns 403 Forbidden to follow standard K8s RBAC behavior
   - Previously returned 404 NotFound for security-by-obscurity
   - Updated test expectations to match new behavior

2. Propagate field and label selectors in Watch handler
   - Pass opts.FieldSelector and opts.LabelSelector to upstream Watch
   - Add defensive filtering before authorization to prevent RBAC bypass
   - Fixes potential issue with resourceNames restrictions

3. Refactor subject-matching logic to eliminate duplication
   - Extract matchesSubject() helper for Group/User/ServiceAccount checks
   - Remove duplicated code from filterAccessible and hasAccessToNamespace
   - Consolidates ServiceAccount namespace fallback logic

All tests pass successfully.

Signed-off-by: IvanHunters <xorokhotnikov@gmail.com>
---
 pkg/registry/core/tenantnamespace/rest.go     | 94 +++++++++++--------
 .../core/tenantnamespace/rest_test.go         |  6 +-
 2 files changed, 56 insertions(+), 44 deletions(-)

diff --git a/pkg/registry/core/tenantnamespace/rest.go b/pkg/registry/core/tenantnamespace/rest.go
index 7724c0e1..223a6925 100644
--- a/pkg/registry/core/tenantnamespace/rest.go
+++ b/pkg/registry/core/tenantnamespace/rest.go
@@ -17,6 +17,8 @@ import (
 	apierrors "k8s.io/apimachinery/pkg/api/errors"
 	metainternal "k8s.io/apimachinery/pkg/apis/meta/internalversion"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/fields"
+	"k8s.io/apimachinery/pkg/labels"
 	"k8s.io/apimachinery/pkg/runtime"
 	"k8s.io/apimachinery/pkg/runtime/schema"
 	"k8s.io/apimachinery/pkg/types"
@@ -130,8 +132,8 @@ func (r *REST) Get(
 		return nil, err
 	}
 	if !hasAccess {
-		// Return NotFound instead of Forbidden to prevent enumeration
-		return nil, apierrors.NewNotFound(r.gvr.GroupResource(), name)
+		// Return Forbidden to follow standard K8s RBAC behavior
+		return nil, apierrors.NewForbidden(r.gvr.GroupResource(), name, fmt.Errorf("access denied"))
 	}
 
 	ns := &corev1.Namespace{}
@@ -155,10 +157,20 @@ func (r *REST) Get(
 
 func (r *REST) Watch(ctx context.Context, opts *metainternal.ListOptions) (watch.Interface, error) {
 	nsList := &corev1.NamespaceList{}
-	nsWatch, err := r.w.Watch(ctx, nsList, &client.ListOptions{Raw: &metav1.ListOptions{
+
+	// Build upstream watch options with field and label selectors
+	rawOpts := &metav1.ListOptions{
 		Watch:           true,
 		ResourceVersion: opts.ResourceVersion,
-	}})
+	}
+	if opts.FieldSelector != nil {
+		rawOpts.FieldSelector = opts.FieldSelector.String()
+	}
+	if opts.LabelSelector != nil {
+		rawOpts.LabelSelector = opts.LabelSelector.String()
+	}
+
+	nsWatch, err := r.w.Watch(ctx, nsList, &client.ListOptions{Raw: rawOpts})
 	if err != nil {
 		return nil, err
 	}
@@ -200,6 +212,18 @@ func (r *REST) Watch(ctx context.Context, opts *metainternal.ListOptions) (watch
 				continue
 			}
 
+			// Apply defensive filtering for field and label selectors
+			if opts.FieldSelector != nil {
+				if !opts.FieldSelector.Matches(fields.Set{"metadata.name": ns.Name}) {
+					continue
+				}
+			}
+			if opts.LabelSelector != nil {
+				if !opts.LabelSelector.Matches(labels.Set(ns.Labels)) {
+					continue
+				}
+			}
+
 			// Check if user has access to this namespace
 			hasAccess, err := r.hasAccessToNamespace(ctx, ns.Name)
 			if err != nil {
@@ -331,6 +355,26 @@ func (r *REST) makeList(src *corev1.NamespaceList, allowed []string) *corev1alph
 	return out
 }
 
+// matchesSubject checks if a RoleBinding subject matches the user's identity.
+// It handles Group, User, and ServiceAccount subjects with proper namespace fallback.
+func matchesSubject(subj rbacv1.Subject, bindingNamespace, username string, groups map[string]struct{}) bool {
+	switch subj.Kind {
+	case "Group":
+		_, ok := groups[subj.Name]
+		return ok
+	case "User":
+		return subj.Name == username
+	case "ServiceAccount":
+		saNamespace := subj.Namespace
+		if saNamespace == "" {
+			saNamespace = bindingNamespace
+		}
+		return username == fmt.Sprintf("system:serviceaccount:%s:%s", saNamespace, subj.Name)
+	default:
+		return false
+	}
+}
+
 func (r *REST) filterAccessible(
 	ctx context.Context,
 	names []string,
@@ -369,26 +413,9 @@ func (r *REST) filterAccessible(
 	subjectLoop:
 		for j := range rbs.Items[i].Subjects {
 			subj := rbs.Items[i].Subjects[j]
-			switch subj.Kind {
-			case "Group":
-				if _, ok = groups[subj.Name]; ok {
-					allowedNameSet[rbs.Items[i].Namespace] = struct{}{}
-					break subjectLoop
-				}
-			case "User":
-				if subj.Name == u.GetName() {
-					allowedNameSet[rbs.Items[i].Namespace] = struct{}{}
-					break subjectLoop
-				}
-			case "ServiceAccount":
-				saNamespace := subj.Namespace
-				if saNamespace == "" {
-					saNamespace = rbs.Items[i].Namespace
-				}
-				if u.GetName() == fmt.Sprintf("system:serviceaccount:%s:%s", saNamespace, subj.Name) {
-					allowedNameSet[rbs.Items[i].Namespace] = struct{}{}
-					break subjectLoop
-				}
+			if matchesSubject(subj, rbs.Items[i].Namespace, u.GetName(), groups) {
+				allowedNameSet[rbs.Items[i].Namespace] = struct{}{}
+				break subjectLoop
 			}
 		}
 	}
@@ -434,23 +461,8 @@ func (r *REST) hasAccessToNamespace(
 	for i := range rbs.Items {
 		for j := range rbs.Items[i].Subjects {
 			subj := rbs.Items[i].Subjects[j]
-			switch subj.Kind {
-			case "Group":
-				if _, ok := groups[subj.Name]; ok {
-					return true, nil
-				}
-			case "User":
-				if subj.Name == u.GetName() {
-					return true, nil
-				}
-			case "ServiceAccount":
-				saNamespace := subj.Namespace
-				if saNamespace == "" {
-					saNamespace = rbs.Items[i].Namespace
-				}
-				if u.GetName() == fmt.Sprintf("system:serviceaccount:%s:%s", saNamespace, subj.Name) {
-					return true, nil
-				}
+			if matchesSubject(subj, rbs.Items[i].Namespace, u.GetName(), groups) {
+				return true, nil
 			}
 		}
 	}
diff --git a/pkg/registry/core/tenantnamespace/rest_test.go b/pkg/registry/core/tenantnamespace/rest_test.go
index eb678949..e1fb1e85 100644
--- a/pkg/registry/core/tenantnamespace/rest_test.go
+++ b/pkg/registry/core/tenantnamespace/rest_test.go
@@ -462,9 +462,9 @@ func TestGet_WithoutAccess(t *testing.T) {
 		t.Errorf("expected nil object, got %v", obj)
 	}
 
-	// Verify it returns NotFound (not Forbidden) to prevent enumeration
-	if !apierrors.IsNotFound(err) {
-		t.Errorf("expected NotFound error, got %v", err)
+	// Verify it returns Forbidden to follow standard K8s RBAC behavior
+	if !apierrors.IsForbidden(err) {
+		t.Errorf("expected Forbidden error, got %v", err)
 	}
 }
 

From 61ed7ad89c0bff66236e6884e6b6cd6038383718 Mon Sep 17 00:00:00 2001
From: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
Date: Tue, 28 Apr 2026 16:33:37 +0500
Subject: [PATCH 82/82] fix(api): address review feedback on TenantNamespace
 watch path

- Hoist user identity extraction out of the Watch goroutine; reuse a
  cached username and groups map across events instead of re-fetching
  them per event. Watch now returns Unauthorized up front when no user
  is present in the context, rather than failing silently per event.
- Switch the per-event access-check error log to structured klog.ErrorS
  to comply with the project Go style guide.
- Strengthen TestGet_WithAccess to assert the concrete *TenantNamespace
  type plus Name, Kind, and APIVersion, so type or metadata regressions
  fail fast.

Assisted-By: Claude <noreply@anthropic.com>
Signed-off-by: Myasnikov Daniil <myasnikovdaniil2001@gmail.com>
---
 pkg/registry/core/tenantnamespace/rest.go     | 35 +++++++++++++++----
 .../core/tenantnamespace/rest_test.go         | 16 +++++++--
 2 files changed, 43 insertions(+), 8 deletions(-)

diff --git a/pkg/registry/core/tenantnamespace/rest.go b/pkg/registry/core/tenantnamespace/rest.go
index 223a6925..4ee97eb0 100644
--- a/pkg/registry/core/tenantnamespace/rest.go
+++ b/pkg/registry/core/tenantnamespace/rest.go
@@ -156,6 +156,18 @@ func (r *REST) Get(
 // -----------------------------------------------------------------------------
 
 func (r *REST) Watch(ctx context.Context, opts *metainternal.ListOptions) (watch.Interface, error) {
+	// Extract user identity once for the lifetime of the watch — it does not
+	// change between events and rebuilding it per event is wasteful.
+	u, ok := request.UserFrom(ctx)
+	if !ok {
+		return nil, apierrors.NewUnauthorized("user missing in context")
+	}
+	username := u.GetName()
+	groups := make(map[string]struct{})
+	for _, group := range u.GetGroups() {
+		groups[group] = struct{}{}
+	}
+
 	nsList := &corev1.NamespaceList{}
 
 	// Build upstream watch options with field and label selectors
@@ -224,10 +236,11 @@ func (r *REST) Watch(ctx context.Context, opts *metainternal.ListOptions) (watch
 				}
 			}
 
-			// Check if user has access to this namespace
-			hasAccess, err := r.hasAccessToNamespace(ctx, ns.Name)
+			// Check if user has access to this namespace using the cached
+			// identity — avoids re-extracting user/groups on every event.
+			hasAccess, err := r.hasAccessToNamespaceForUser(ctx, ns.Name, username, groups)
 			if err != nil {
-				klog.Errorf("Failed to check access for namespace %s in watch: %v", ns.Name, err)
+				klog.ErrorS(err, "Failed to check access for namespace in watch", "namespace", ns.Name)
 				continue
 			}
 			if !hasAccess {
@@ -437,12 +450,22 @@ func (r *REST) hasAccessToNamespace(
 	if !ok {
 		return false, fmt.Errorf("user missing in context")
 	}
-
-	// Check privileged groups
 	groups := make(map[string]struct{})
 	for _, group := range u.GetGroups() {
 		groups[group] = struct{}{}
 	}
+	return r.hasAccessToNamespaceForUser(ctx, namespace, u.GetName(), groups)
+}
+
+// hasAccessToNamespaceForUser is the inner check that does not re-extract user
+// identity from context. Use this in hot paths (e.g. the Watch loop) where the
+// caller has already cached the user name and groups.
+func (r *REST) hasAccessToNamespaceForUser(
+	ctx context.Context,
+	namespace, username string,
+	groups map[string]struct{},
+) (bool, error) {
+	// Check privileged groups
 	if _, ok := groups["system:masters"]; ok {
 		return true, nil
 	}
@@ -461,7 +484,7 @@ func (r *REST) hasAccessToNamespace(
 	for i := range rbs.Items {
 		for j := range rbs.Items[i].Subjects {
 			subj := rbs.Items[i].Subjects[j]
-			if matchesSubject(subj, rbs.Items[i].Namespace, u.GetName(), groups) {
+			if matchesSubject(subj, rbs.Items[i].Namespace, username, groups) {
 				return true, nil
 			}
 		}
diff --git a/pkg/registry/core/tenantnamespace/rest_test.go b/pkg/registry/core/tenantnamespace/rest_test.go
index e1fb1e85..52d2a075 100644
--- a/pkg/registry/core/tenantnamespace/rest_test.go
+++ b/pkg/registry/core/tenantnamespace/rest_test.go
@@ -15,6 +15,8 @@ import (
 	"k8s.io/apiserver/pkg/authentication/user"
 	"k8s.io/apiserver/pkg/endpoints/request"
 	"sigs.k8s.io/controller-runtime/pkg/client/fake"
+
+	corev1alpha1 "github.com/cozystack/cozystack/pkg/apis/core/v1alpha1"
 )
 
 func TestMakeListSortsAlphabetically(t *testing.T) {
@@ -408,8 +410,18 @@ func TestGet_WithAccess(t *testing.T) {
 	if err != nil {
 		t.Fatalf("unexpected error: %v", err)
 	}
-	if obj == nil {
-		t.Fatal("expected object, got nil")
+	tn, ok := obj.(*corev1alpha1.TenantNamespace)
+	if !ok {
+		t.Fatalf("expected *TenantNamespace, got %T", obj)
+	}
+	if tn.Name != "tenant-test" {
+		t.Errorf("expected name %q, got %q", "tenant-test", tn.Name)
+	}
+	if tn.Kind != "TenantNamespace" {
+		t.Errorf("expected Kind=TenantNamespace, got %q", tn.Kind)
+	}
+	if tn.APIVersion != corev1alpha1.SchemeGroupVersion.String() {
+		t.Errorf("expected APIVersion=%q, got %q", corev1alpha1.SchemeGroupVersion.String(), tn.APIVersion)
 	}
 }