monitoring: surface CrashLoopBackOff pods as degraded on the frontend

monitorFrontendStatus's "pod" branch was re-deriving status from
Kubernetes.PodPhase alone, so a CrashLoopBackOff pod (Phase=Running
with unready containers) was reported as "running" even though
statusFromKubernetesPod had already classified it as StatusWarning.

Drop the unconditional phase=running -> "running" mapping and fall
through to the unified-status switch, which surfaces StatusWarning as
"degraded". Phase-based mappings for pending/unknown/succeeded/failed
are kept since they still cover edge cases where the unified status
is StatusOnline but phase carries useful information.

Adds two regression cases in TestMonitorFrontendAndMetricHelpers:
- pod warning running degraded (CrashLoopBackOff)
- pod online running running (healthy baseline)

internal/monitoring/guest_config_conversion_coverage_test.go

@@ -437,6 +437,29 @@ func TestMonitorFrontendAndMetricHelpers(t *testing.T) {
 				},
 				want: "stopped",
 			},
+			{
+				// CrashLoopBackOff: phase stays "Running" but containers are
+				// unready, so statusFromKubernetesPod returns StatusWarning.
+				// The frontend mapping must surface that as "degraded" rather
+				// than reporting the pod as healthy on the strength of phase
+				// alone.
+				name:         "pod warning running degraded",
+				resourceType: "pod",
+				resource: unifiedresources.Resource{
+					Status:     unifiedresources.StatusWarning,
+					Kubernetes: &unifiedresources.K8sData{PodPhase: "running"},
+				},
+				want: "degraded",
+			},
+			{
+				name:         "pod online running running",
+				resourceType: "pod",
+				resource: unifiedresources.Resource{
+					Status:     unifiedresources.StatusOnline,
+					Kubernetes: &unifiedresources.K8sData{PodPhase: "running"},
+				},
+				want: "running",
+			},
 			{
 				name:         "workload online running fallback",
 				resourceType: "system-container",

internal/monitoring/monitor.go

@@ -5667,13 +5667,15 @@ func monitorFrontendStatus(resource unifiedresources.Resource, resourceType stri
 		if resource.Kubernetes != nil {
 			phase := strings.ToLower(strings.TrimSpace(resource.Kubernetes.PodPhase))
 			switch phase {
-			case "running":
-				return "running"
 			case "pending", "unknown":
 				return "degraded"
 			case "succeeded", "failed":
 				return "stopped"
 			}
+			// Phase=Running is not enough — a pod can have Phase=Running with
+			// CrashLoopBackOff containers, which statusFromKubernetesPod
+			// correctly classifies as StatusWarning. Fall through to the
+			// unified-status mapping below so that surfaces as "degraded".
 		}
 	}