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. 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/hack/admin-kubeconfig-invariant.bats b/hack/admin-kubeconfig-invariant.bats new file mode 100644 index 00000000..a4b98b00 --- /dev/null +++ b/hack/admin-kubeconfig-invariant.bats @@ -0,0 +1,145 @@ +#!/usr/bin/env bats +# ----------------------------------------------------------------------------- +# Chart-wide invariant for packages/apps/kubernetes: +# +# Every Deployment in this chart that mounts -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 $matched Deployment(s)" +} + +@test "chart emits zero admin-kubeconfig Deployments when tenant has no etcd DataStore" { + # Without a DataStore (parent Tenant has not populated _namespace.etcd yet) + # the control-plane-side Deployments must NOT render at all. If they did, + # the wait-for-kubeconfig init would CrashLoopBackOff indefinitely - there + # would be no KamajiControlPlane to provision the Secret - consuming the + # HelmRelease wait budget and triggering exactly the remediation cycle the + # rest of this chart tries to avoid. This test renders the whole chart + # with etcd empty and asserts no Deployment references the admin-kubeconfig + # Secret. + tmp=$(mktemp -d) + trap 'rm -rf "$tmp"' EXIT + + helm template invariant packages/apps/kubernetes \ + --namespace tenant-root \ + --values packages/apps/kubernetes/tests/values-ci-no-etcd.yaml \ + 2>/dev/null > "$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)" +} + +@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 \ + --values packages/apps/kubernetes/tests/values-ci-no-etcd.yaml \ + 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/hack/e2e-apps/remediation-guard.sh b/hack/e2e-apps/remediation-guard.sh new file mode 100644 index 00000000..b45561b9 --- /dev/null +++ b/hack/e2e-apps/remediation-guard.sh @@ -0,0 +1,39 @@ +# Helpers for asserting that a Flux HelmRelease did not fall into an +# install/upgrade remediation cycle during an e2e run. +# +# 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. +# +# 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() { + statuses="$1" + if [ -z "${statuses}" ]; then + return 1 + fi + # 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 be6dcd6f..761ee57d 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="$2" @@ -320,6 +322,35 @@ EOF done kubectl wait hr kubernetes-${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. + # 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}') + # 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' "${history_statuses:-}" + if [ -z "${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." >&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-${test_name}" diff --git a/hack/remediation-guard.bats b/hack/remediation-guard.bats new file mode 100644 index 00000000..092fe06d --- /dev/null +++ b/hack/remediation-guard.bats @@ -0,0 +1,127 @@ +#!/usr/bin/env bats +# ----------------------------------------------------------------------------- +# Unit tests for hack/e2e-apps/remediation-guard.sh +# +# helmrelease_has_remediation_cycle takes a newline-delimited list of +# HelmRelease history snapshot status values (deployed/superseded/failed/ +# uninstalled/...) and returns 0 when any entry is "failed" or "uninstalled" +# (meaning flux helm-controller performed install/upgrade remediation). +# +# This is used by the e2e script after the HelmRelease reaches Ready. The +# failure/upgrade counters (.status.installFailures / .status.upgradeFailures) +# are useless there because flux's ClearFailures zeroes them on successful +# reconciliation; .status.history retains the snapshot trail. +# +# cozytest.sh's awk parser recognizes only @test blocks and a bare `}` on +# its own line; there is no bats `run` or `$status`. Assertions are +# expressed as direct shell tests that exit non-zero on failure. +# +# Run with: hack/cozytest.sh hack/remediation-guard.bats +# ----------------------------------------------------------------------------- + +@test "empty history returns not-detected" { + . hack/e2e-apps/remediation-guard.sh + if helmrelease_has_remediation_cycle ""; then + echo "expected not-detected for empty history" >&2 + exit 1 + fi +} + +@test "single deployed snapshot returns not-detected" { + . hack/e2e-apps/remediation-guard.sh + if helmrelease_has_remediation_cycle "deployed"; then + echo "expected not-detected for deployed-only history" >&2 + exit 1 + fi +} + +@test "deployed then superseded returns not-detected" { + . hack/e2e-apps/remediation-guard.sh + 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 "single failed snapshot returns detected" { + . hack/e2e-apps/remediation-guard.sh + if ! helmrelease_has_remediation_cycle "failed"; then + echo "expected detected when history contains failed snapshot" >&2 + exit 1 + fi +} + +@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 + if ! helmrelease_has_remediation_cycle "uninstalled"; then + echo "expected detected when history contains uninstalled snapshot" >&2 + exit 1 + fi +} + +@test "uninstalled then deployed still returns detected" { + . hack/e2e-apps/remediation-guard.sh + statuses=$(printf 'uninstalled\ndeployed\n') + if ! helmrelease_has_remediation_cycle "${statuses}"; then + echo "expected detected despite later successful deploy" >&2 + exit 1 + fi +} + +@test "deployed then failed still returns detected" { + . hack/e2e-apps/remediation-guard.sh + 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 "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, + # 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 + + cat > "$tmp/hr.yaml" <<'YAML' +apiVersion: helm.toolkit.fluxcd.io/v2 +kind: HelmRelease +metadata: + name: kubernetes-test + namespace: tenant-test +status: + history: + - name: kubernetes-test + namespace: tenant-test + version: 1 + status: uninstalled + - name: kubernetes-test + namespace: tenant-test + version: 2 + status: deployed +YAML + + # Default yq output is yaml scalar format, which for string values emits + # bare unquoted tokens - matching what kubectl -o jsonpath produces in + # e2e. Do not switch to JSON output here; that would quote the values + # and break the loop in helmrelease_has_remediation_cycle. + statuses=$(yq '.status.history[].status' "$tmp/hr.yaml") + + [ -n "$statuses" ] + echo "$statuses" | grep --quiet '^uninstalled$' + + . hack/e2e-apps/remediation-guard.sh + if ! helmrelease_has_remediation_cycle "$statuses"; then + echo "expected detected for pinned HR snippet with uninstalled + deployed history" >&2 + exit 1 + fi +} diff --git a/packages/apps/kubernetes/Makefile b/packages/apps/kubernetes/Makefile index 01cf736d..4fea42ef 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 @@ -67,3 +70,4 @@ 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 + 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/images/busybox.tag b/packages/apps/kubernetes/images/busybox.tag new file mode 100644 index 00000000..39de220a --- /dev/null +++ b/packages/apps/kubernetes/images/busybox.tag @@ -0,0 +1 @@ +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 36c06b64..40a8ae7a 100644 --- a/packages/apps/kubernetes/templates/_helpers.tpl +++ b/packages/apps/kubernetes/templates/_helpers.tpl @@ -49,3 +49,52 @@ 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 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 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 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: "{{ default (.Files.Get "images/busybox.tag" | trim) .Values.images.waitForKubeconfig }}" + 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 a00e0155..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 @@ -23,6 +34,8 @@ spec: - key: node-role.kubernetes.io/control-plane operator: Exists effect: "NoSchedule" + initContainers: + {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }} containers: - image: "{{ $.Files.Get "images/cluster-autoscaler.tag" | trim }}" name: cluster-autoscaler @@ -56,6 +69,7 @@ spec: name: cloud-config - secret: secretName: {{ .Release.Name }}-admin-kubeconfig + optional: true name: kubeconfig serviceAccountName: {{ .Release.Name }}-cluster-autoscaler terminationGracePeriodSeconds: 10 @@ -105,3 +119,4 @@ rules: - list - update - watch +{{- end }} diff --git a/packages/apps/kubernetes/templates/cluster.yaml b/packages/apps/kubernetes/templates/cluster.yaml index 10d6fd80..e1494e64 100644 --- a/packages/apps/kubernetes/templates/cluster.yaml +++ b/packages/apps/kubernetes/templates/cluster.yaml @@ -1,4 +1,15 @@ {{- $etcd := .Values._namespace.etcd }} +{{- /* + 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 }} @@ -84,6 +95,26 @@ 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 }} +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 @@ -404,3 +435,4 @@ metadata: spec: {{- .spec | toYaml | nindent 2 }} {{- end }} +{{- end }} diff --git a/packages/apps/kubernetes/templates/csi/deploy.yaml b/packages/apps/kubernetes/templates/csi/deploy.yaml index 938b6d67..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: @@ -24,6 +25,8 @@ spec: - key: node-role.kubernetes.io/control-plane operator: Exists effect: "NoSchedule" + initContainers: + {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }} containers: - name: csi-driver imagePullPolicy: Always @@ -234,4 +237,6 @@ spec: emptyDir: {} - secret: secretName: {{ .Release.Name }}-admin-kubeconfig + optional: true name: kubeconfig +{{- end }} 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 }} diff --git a/packages/apps/kubernetes/templates/kccm/manager.yaml b/packages/apps/kubernetes/templates/kccm/manager.yaml index 81426d4e..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: @@ -22,6 +23,8 @@ spec: - key: node-role.kubernetes.io/control-plane operator: Exists effect: "NoSchedule" + initContainers: + {{- include "kubernetes.waitForAdminKubeconfig" $ | nindent 6 }} containers: - name: kubevirt-cloud-controller-manager args: @@ -55,5 +58,7 @@ spec: name: cloud-config - secret: secretName: {{ .Release.Name }}-admin-kubeconfig + optional: true name: kubeconfig serviceAccountName: {{ .Release.Name }}-kccm +{{- end }} 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..507e00af --- /dev/null +++ b/packages/apps/kubernetes/tests/admin_kubeconfig_wait_test.yaml @@ -0,0 +1,155 @@ +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 + 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: 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 + 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 + 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: 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: + path: kind + value: KamajiControlPlane + asserts: + - equal: + path: spec.dataStoreName + value: tenant-root + + - 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: + etcd: "" + monitoring: "" + ingress: "" + seaweedfs: "" + host: "" + asserts: + - hasDocuments: + count: 1 + - isKind: + of: ConfigMap + - equal: + path: metadata.name + value: test-awaiting-etcd + - equal: + path: data.status + value: awaiting-etcd 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/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 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 5e9e8f94..1631c86e 100644 --- a/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml +++ b/packages/system/kubernetes-rd/cozyrds/kubernetes.yaml @@ -2,13 +2,27 @@ 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 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: kind: Kubernetes 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 `.` 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 `.` 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: @@ -26,7 +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: diff --git a/pkg/cmd/server/start.go b/pkg/cmd/server/start.go index f7791332..7ebbbc89 100644 --- a/pkg/cmd/server/start.go +++ b/pkg/cmd/server/start.go @@ -160,6 +160,34 @@ 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. 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, @@ -168,15 +196,7 @@ 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..21de27b2 100644 --- a/pkg/config/config.go +++ b/pkg/config/config.go @@ -16,6 +16,48 @@ limitations under the License. package config +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 { Resources []Resource `yaml:"resources"` @@ -41,6 +83,12 @@ 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/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) + } + }) + } +} diff --git a/pkg/registry/apps/application/rest.go b/pkg/registry/apps/application/rest.go index 0728ea13..d233ecc3 100644 --- a/pkg/registry/apps/application/rest.go +++ b/pkg/registry/apps/application/rest.go @@ -1528,6 +1528,24 @@ func (r *REST) convertApplicationToHelmRelease(app *appsv1alpha1.Application) (* }, } + // 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 + 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 new file mode 100644 index 00000000..85ac0b97 --- /dev/null +++ b/pkg/registry/apps/application/rest_timeout_test.go @@ -0,0 +1,116 @@ +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 newRESTForTimeout(kind, prefix string, helmInstallTimeout time.Duration) *REST { + return &REST{ + kindName: kind, + releaseConfig: config.ReleaseConfig{ + Prefix: prefix, + ChartRef: config.ChartRefConfig{ + Kind: "HelmChart", + Name: "x", + Namespace: "cozy-system", + }, + HelmInstallTimeout: helmInstallTimeout, + }, + } +} + +// 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, + }, + { + // 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, + }, + { + name: "arbitrary future kind with 20m configured gets 20m", + kind: "TalosCluster", + prefix: "talos-", + configured: 20 * time.Minute, + wantSet: true, + }, + } + + 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"}, + } + + hr, err := r.convertApplicationToHelmRelease(app) + if err != nil { + t.Fatalf("convertApplicationToHelmRelease returned error: %v", err) + } + + if hr.Spec.Install == nil || hr.Spec.Upgrade == nil { + t.Fatalf("Spec.Install/Upgrade must be non-nil") + } + + 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) + } + }) + } +}