Fix waiting and erroring out on garbage-collected Jobs

.github/workflows/e2e.yaml

@@ -80,6 +80,14 @@ jobs:
           kubectl -n helm-system wait helmreleases/podinfo-git --for=condition=ready --timeout=4m
           kubectl -n helm-system wait helmreleases/podinfo-oci --for=condition=ready --timeout=4m
           kubectl -n helm-system delete -f config/testdata/podinfo
+      - name: Run Job with TTL test
+        run: |
+          # This test verifies that the wait logic correctly handles Jobs with
+          # ttlSecondsAfterFinished that get garbage-collected after completion.
+          # Without the fix, the wait would fail with NotFound error.
+          kubectl -n helm-system apply -f config/testdata/job-ttl
+          kubectl -n helm-system wait helmreleases/job-ttl --for=condition=ready --timeout=4m
+          kubectl -n helm-system delete -f config/testdata/job-ttl
       - name: Run client-side apply upgrade test
         run: |
           set -euo pipefail
@@ -1147,7 +1155,7 @@ jobs:
             exit 1
           fi
           kubectl -n helm-system delete -f config/testdata/post-renderer-kustomize
-      - name: Bootstrap CRDs Upgrade Tests
+      - name: Bootstrap Tests Using Local Helm Chart
         if: ${{ startsWith(github.ref, 'refs/tags/') || startsWith(github.ref, 'refs/heads/') }}
         run: |
           REF=${{ github.ref }}

config/testdata/job-ttl/helmrelease.yaml

@@ -0,0 +1,25 @@
+apiVersion: helm.toolkit.fluxcd.io/v2
+kind: HelmRelease
+metadata:
+  name: job-ttl
+spec:
+  interval: 5m
+  timeout: 2m
+  chart:
+    spec:
+      chart: podinfo
+      version: '>=6.10.1 <7.0.0'
+      sourceRef:
+        kind: HelmRepository
+        name: podinfo
+      interval: 1m
+  values:
+    # Enable a post-install hook Job with ttlSecondsAfterFinished.
+    # The Job completes immediately and gets garbage-collected by the TTL controller.
+    # The fix ensures the wait doesn't fail when the Job is NotFound after being
+    # garbage-collected.
+    hooks:
+      postInstall:
+        job:
+          enabled: true
+          ttlSecondsAfterFinished: 0

go.mod

@@ -26,7 +26,7 @@ require (
 	github.com/fluxcd/pkg/cache v0.13.0
 	github.com/fluxcd/pkg/chartutil v1.21.0
 	github.com/fluxcd/pkg/runtime v0.97.0
-	github.com/fluxcd/pkg/ssa v0.65.0
+	github.com/fluxcd/pkg/ssa v0.66.0
 	github.com/fluxcd/pkg/testserver v0.13.0
 	github.com/fluxcd/source-controller/api v1.7.2
 	github.com/go-logr/logr v1.4.3

go.sum

@@ -162,8 +162,8 @@ github.com/fluxcd/pkg/chartutil v1.21.0 h1:NJYhlekwBwuqMpRgsOlcsJrw2Xq0cBJW0Nmvz
 github.com/fluxcd/pkg/chartutil v1.21.0/go.mod h1:Gv50bF3SS4OvvKCyyIMRkGeNzZk6Fsh4+lAdrjx97T4=
 github.com/fluxcd/pkg/runtime v0.97.0 h1:NZ9ZL2RE6J4yNDHI7ZNmL/OnCsDWurzVIvws73FYMz8=
 github.com/fluxcd/pkg/runtime v0.97.0/go.mod h1:FyjNMFNAERkCsF/muTWJYU9MZOsq/m4Sc4aQk/EgQ9E=
-github.com/fluxcd/pkg/ssa v0.65.0 h1:pvPYmBmfqS0sRNoi4/IUoocBbglYuD/OJLgOuxMJ9+o=
-github.com/fluxcd/pkg/ssa v0.65.0/go.mod h1:RjvVjJIoRo1ecsv91yMuiqzO6cpNag80M6MOB/vrJdc=
+github.com/fluxcd/pkg/ssa v0.66.0 h1:CYbCTuws8Sn1xAaOFNz92yS2iS5YBqJgd+DrzmRxfaQ=
+github.com/fluxcd/pkg/ssa v0.66.0/go.mod h1:RjvVjJIoRo1ecsv91yMuiqzO6cpNag80M6MOB/vrJdc=
 github.com/fluxcd/pkg/testserver v0.13.0 h1:xEpBcEYtD7bwvZ+i0ZmChxKkDo/wfQEV3xmnzVybSSg=
 github.com/fluxcd/pkg/testserver v0.13.0/go.mod h1:akRYv3FLQUsme15na9ihECRG6hBuqni4XEY9W8kzs8E=
 github.com/fluxcd/source-controller/api v1.7.2 h1:/lg/xoyRjxwdhHKqjTxQS2o1cp+DMKJ8W4rpm+ZLemQ=

internal/action/wait.go

@@ -25,14 +25,16 @@ import (
 	"github.com/fluxcd/cli-utils/pkg/kstatus/status"
 	"github.com/fluxcd/cli-utils/pkg/object"
 	helmkube "helm.sh/helm/v4/pkg/kube"
+	appsv1 "k8s.io/api/apps/v1"
+	batchv1 "k8s.io/api/batch/v1"
 	apimeta "k8s.io/apimachinery/pkg/api/meta"
 	"k8s.io/apimachinery/pkg/apis/meta/v1/unstructured"
 	"k8s.io/apimachinery/pkg/runtime"
-	"k8s.io/apimachinery/pkg/runtime/schema"
 
 	"github.com/fluxcd/pkg/runtime/controller"
 	runtimestatusreaders "github.com/fluxcd/pkg/runtime/statusreaders"
 	"github.com/fluxcd/pkg/ssa"
+	ssautils "github.com/fluxcd/pkg/ssa/utils"
 
 	v2 "github.com/fluxcd/helm-controller/api/v2"
 )
@@ -152,10 +154,12 @@ func (w *waiter) wait(ctx context.Context, resources helmkube.ResourceList,
 	timeout time.Duration, failFast bool, sr ...NewStatusReaderFunc) error {
 
 	// WaitForSetWithContext expects a list of ObjMetadata.
-	objs := []object.ObjMetadata{}
+	var objs object.ObjMetadataSet
+	var jobs []*unstructured.Unstructured
 	for _, res := range resources {
-		// Skip paused apps/v1/Deployment (copied from Helm).
 		gvk := res.Object.GetObjectKind().GroupVersionKind()
+
+		// Skip paused apps/v1/Deployment (copied from Helm).
 		if gvk == deploymentGVK {
 			uns, err := runtime.DefaultUnstructuredConverter.ToUnstructured(res.Object)
 			if err != nil {
@@ -167,6 +171,15 @@ func (w *waiter) wait(ctx context.Context, resources helmkube.ResourceList,
 			}
 		}
 
+		// Collect Jobs with TTL for special handling.
+		if gvk == jobGVK {
+			uns, err := runtime.DefaultUnstructuredConverter.ToUnstructured(res.Object)
+			if err != nil {
+				return err
+			}
+			jobs = append(jobs, &unstructured.Unstructured{Object: uns})
+		}
+
 		// Convert to ObjMetadata.
 		obj, err := object.RuntimeToObjMeta(res.Object)
 		if err != nil {
@@ -176,12 +189,16 @@ func (w *waiter) wait(ctx context.Context, resources helmkube.ResourceList,
 	}
 
 	return w.newResourceManager(sr...).WaitForSetWithContext(ctx, objs, ssa.WaitOptions{
-		Interval: 5 * time.Second, // Copied from kustomize-controller.
-		Timeout:  timeout,
+		JobsWithTTL: ssautils.ExtractJobsWithTTL(jobs),
+		Interval:    5 * time.Second, // Copied from kustomize-controller.
+		Timeout:     timeout,
 		// The kustomize-controller has an opt-in feature gate that disables
 		// fail fast here: DisableFailFastBehavior.
 		FailFast: failFast,
 	})
 }
 
-var deploymentGVK = schema.GroupVersionKind{Group: "apps", Version: "v1", Kind: "Deployment"}
+var (
+	deploymentGVK = appsv1.SchemeGroupVersion.WithKind("Deployment")
+	jobGVK        = batchv1.SchemeGroupVersion.WithKind("Job")
+)