When e2e failover cmd was used to failover cephfs appset workload, the cmd output shows passed as seen below:
./run.sh -test.run TestDR/appset-deploy-cephfs/Failover
2026-08-06T15:34:34.864+0530 INFO Using config file "config.yaml"
2026-08-06T15:34:34.865+0530 INFO Using log file "ramen-e2e.log"
2026-08-06T15:34:34.871+0530 INFO Using "hub" cluster name: "hub"
2026-08-06T15:34:34.928+0530 INFO Detected "c1" managed cluster name: "dr1"
2026-08-06T15:34:34.984+0530 INFO Detected "c2" managed cluster name: "dr2"
2026-08-06T15:34:34.985+0530 INFO Using DeployTimeout: 5m0s
2026-08-06T15:34:34.985+0530 INFO Using UneployTimeout: 5m0s
2026-08-06T15:34:34.985+0530 INFO Using EnableTimeout: 5m0s
2026-08-06T15:34:34.985+0530 INFO Using DisableTimeout: 10m0s
2026-08-06T15:34:34.985+0530 INFO Using FailoverTimeout: 15m0s
2026-08-06T15:34:34.985+0530 INFO Using RelocateTimeout: 15m0s
2026-08-06T15:34:34.985+0530 INFO Using RetryInterval: 5s
=== RUN TestDR
=== PAUSE TestDR
=== CONT TestDR
2026-08-06T15:34:35.002+0530 INFO Detected kubernetes distribution: "k8s"
2026-08-06T15:34:35.002+0530 INFO Using namespaces: {RamenHubNamespace:ramen-system RamenDRClusterNamespace:ramen-system RamenOpsNamespace:ramen-ops ArgocdNamespace:argocd}
2026-08-06T15:34:35.050+0530 INFO Validated clusters ["dr1", "dr2"] in DRPolicy "dr-policy-1m"
2026-08-06T15:34:35.060+0530 INFO Validated clusters ["dr1", "dr2"] in ClusterSet "default"
2026-08-06T15:34:35.090+0530 INFO Created channel "test-gitops/https-github-com-ramendr-ocm-ramen-samples-git" in cluster "hub"
=== RUN TestDR/appset-deploy-cephfs
=== PAUSE TestDR/appset-deploy-cephfs
=== CONT TestDR/appset-deploy-cephfs
=== RUN TestDR/appset-deploy-cephfs/Failover
2026-08-06T15:34:35.171+0530 INFO appset-deploy-cephfs Failing over workload from cluster "dr2" to cluster "dr1"
2026-08-06T15:37:51.065+0530 INFO appset-deploy-cephfs Workload failed over
2026-08-06T15:37:51.081+0530 INFO Deleted channel "test-gitops/https-github-com-ramendr-ocm-ramen-samples-git" in cluster "hub"
--- PASS: TestDR (6.19s)
--- PASS: TestDR/appset-deploy-cephfs (195.97s)
--- PASS: TestDR/appset-deploy-cephfs/Failover (195.97s)
But when checked locally, workload was still running on older primary (dr2) and VRG CURRENTSTATE was Unknown.
We should ensure that VRG has successfully transitioned to secondary and matches the desired state.
oc get deploy,pods,pvc,vr,vrg,vgr -n test-appset-deploy-cephfs --context dr2
NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/busybox 1/1 1 1 11m
NAME READY STATUS RESTARTS AGE
pod/busybox-54d5b75d6c-tf4mt 1/1 Running 0 11m
pod/volsync-rsync-tls-src-busybox-pvc-dt5hd 0/1 Error 0 72s
pod/volsync-rsync-tls-src-busybox-pvc-pml8t 0/1 Error 0 2m30s
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS VOLUMEATTRIBUTESCLASS AGE
persistentvolumeclaim/busybox-pvc Bound pvc-1722fd50-d95a-461d-b978-656cf961583b 1Gi RWX rook-cephfs-fs1 <unset> 17m
persistentvolumeclaim/volsync-busybox-pvc-src Bound pvc-8a5be97e-705a-4f22-8f9c-a763ed5d0620 1Gi RWX rook-cephfs-fs1 <unset> 5m23s
NAME DESIREDSTATE CURRENTSTATE
volumereplicationgroup.ramendr.openshift.io/appset-deploy-cephfs secondary Unknown
This is not a timing issue as the cleanup took further 3-5mins to complete after the test was marked as passed.
May be (not sure) this happens for relocate as well, but this issue was found during failover operation, where the same workload was failedover multiple times.
When e2e failover cmd was used to failover cephfs appset workload, the cmd output shows passed as seen below:
But when checked locally, workload was still running on older primary (dr2) and VRG CURRENTSTATE was Unknown.
We should ensure that VRG has successfully transitioned to secondary and matches the desired state.
This is not a timing issue as the cleanup took further 3-5mins to complete after the test was marked as passed.
May be (not sure) this happens for relocate as well, but this issue was found during failover operation, where the same workload was failedover multiple times.