fix: PVC in wf.status should be reset when retrying workflow (#10685)

Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>

test/e2e/cli_test.go

@@ -890,6 +890,30 @@ func (s *CLISuite) TestWorkflowRetryNestedDag() {
  })
 }
 
+func (s *CLISuite) TestWorkflowRetryWithRecreatedPVC() {
+ s.Given().
+ Workflow("@testdata/retry-with-recreated-pvc-test.yaml").
+ When().
+ SubmitWorkflow().
+ WaitForWorkflow(fixtures.ToBeFailed).
+ Then().
+ RunCli([]string{"retry", "retry-with-recreated-pvc"}, func(t *testing.T, output string, err error) {
+ if assert.NoError(t, err, output) {
+ assert.Contains(t, output, "Name:")
+ assert.Contains(t, output, "Namespace:")
+ }
+ }).
+ When().
+ WaitForWorkflow(fixtures.ToBeFailed).
+ Then().
+ ExpectWorkflow(func(t *testing.T, _ *metav1.ObjectMeta, status *wfv1.WorkflowStatus) {
+ assert.Equal(t, wfv1.NodeFailed, status.Nodes.FindByDisplayName("print").Phase)
+ // This step is failed intentionally to allow retry. The error message is not related to PVC that is deleted
+ // previously since it is re-created during retry.
+ assert.Equal(t, "Error (exit code 1)", status.Nodes.FindByDisplayName("print").Message)
+ })
+}
+
 func (s *CLISuite) TestWorkflowStop() {
  s.Given().
  Workflow("@smoke/basic.yaml").

test/e2e/testdata/retry-with-recreated-pvc-test.yaml

@@ -0,0 +1,42 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Workflow
+metadata:
+ name: retry-with-recreated-pvc
+spec:
+ volumeClaimGC:
+ strategy: OnWorkflowCompletion
+ entrypoint: volumes-pvc-example
+ volumeClaimTemplates:
+ - metadata:
+ name: workdir
+ spec:
+ accessModes: [ "ReadWriteOnce" ]
+ resources:
+ requests:
+ storage: 1Gi
+
+ templates:
+ - name: volumes-pvc-example
+ steps:
+ - - name: generate
+ template: whalesay
+ - - name: print
+ template: print-message
+
+ - name: whalesay
+ container:
+ image: argoproj/argosay:v2
+ command: [sh, -c]
+ args: ["echo generating message in volume; cowsay hello world | tee /mnt/vol/hello_world.txt"]
+ volumeMounts:
+ - name: workdir
+ mountPath: /mnt/vol
+
+ - name: print-message
+ container:
+ image: argoproj/argosay:v2
+ command: [sh, -c]
+ args: ["exit 1"]
+ volumeMounts:
+ - name: workdir
+ mountPath: /mnt/vol

workflow/util/util.go

@@ -821,6 +821,7 @@ func FormulateRetryWorkflow(ctx context.Context, wf *wfv1.Workflow, restartSucce
  newWF.Status.StartedAt = metav1.Time{Time: time.Now().UTC()}
  newWF.Status.FinishedAt = metav1.Time{}
  newWF.Spec.Shutdown = ""
+ newWF.Status.PersistentVolumeClaims = []apiv1.Volume{}
  if newWF.Spec.ActiveDeadlineSeconds != nil && *newWF.Spec.ActiveDeadlineSeconds == 0 {
  // if it was terminated, unset the deadline
  newWF.Spec.ActiveDeadlineSeconds = nil