Bug 2283038 - [RDR] [Hub recovery] [Co-situated] Failover never completes when peer ready is true but replication destination is missing [NEEDINFO]
Summary: [RDR] [Hub recovery] [Co-situated] Failover never completes when peer ready i...
Keywords:
Status: ASSIGNED
Alias: None
Product: Red Hat OpenShift Data Foundation
Classification: Red Hat Storage
Component: odf-dr
Version: 4.16
Hardware: Unspecified
OS: Unspecified
unspecified
high
Target Milestone: ---
: ---
Assignee: Benamar Mekhissi
QA Contact: krishnaram Karthick
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2024-05-23 17:51 UTC by Aman Agrawal
Modified: 2024-09-26 09:45 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Known Issue
Doc Text:
.Failover process fails when the `ReplicationDestination` resource has not been created yet If the user initiates a failover before the `LastGroupSyncTime` is updated, the failover process might fail. This failure is accompanied by an error message indicating that the `ReplicationDestination` does not exist. Workaround: To mitigate this issue, follow these steps: . Edit the `ManifestWork` for the VRG on the hub cluster. . Delete the following section from the manifest: ``` /spec/workload/manifests/0/spec/volsync ``` . Save the changes. Applying this workaround correctly ensures that the VRG skips attempting to restore the PVC using the `ReplicationDestination` resource. If the PVC already exists, the application uses it as is. If the PVC does not exist, a new PVC is created.
Clone Of:
Environment:
Last Closed:
Embargoed:
kseeger: needinfo?
amagrawa: needinfo? (bmekhiss)
amagrawa: needinfo? (edonnell)

Attachments (Terms of Use)

Description Aman Agrawal 2024-05-23 17:51:02 UTC 
Description of problem (please be detailed as possible and provide log
snippests):


Version of all relevant components (if applicable):
ceph version 18.2.1-136.el9cp (e7edde2b655d0dd9f860dda675f9d7954f07e6e3) reef (stable)
OCP 4.16.0-0.nightly-2024-04-26-145258
ODF 4.16.0-89.stable
ACM 2.10.2
MCE 2.5.2



Does this issue impact your ability to continue to work with the product
(please explain in detail what is the user impact)?


Is there any workaround available to the best of your knowledge?


Rate from 1 - 5 the complexity of the scenario you performed that caused this
bug (1 - very simple, 5 - very complex)?


Can this issue reproducible?


Can this issue reproduce from the UI?


If this is a regression, please provide more details to justify this:


Steps to Reproduce:

*****Active hub co-situated with primary managed cluster*****

1. On a RDR setup with both RBD and CephFS workloads of subscription and appset (pull model) types in distinct states like Deployed, FailedOver and Relocated, perform site-failure by bringing active hub and primary managed cluster down and move to passive hub by performing hub recovery.
2. Then failover all the workloads running on down managed cluster to the surviving managed cluster.
3. After successful failover, recover the down managed cluster.

During cleanup, VRG both states would be marked as Secondary for cephfs workloads on the recovered managed cluster which would eventually mark peer ready as True in the drpc resource on hub but the replicationdestination would not be created on the recovered cluster until the eviction period timeout which is 24hrs as of now. 

4. Now failover cephfs workloads back to the surviving cluster where peer ready is marked as true but replication destination isn't created.


Actual results: Since peer ready is marked as true for cephfs workloads in this case, UI will allow failover even if 1st sync has not completed due to missing replication destination.

Marking peer ready is expected when VRG both states are marked as secondary on the recovered cluster (refer comment https://bugzilla.redhat.com/show_bug.cgi?id=2263488#c21), failover never completes when replication destination is missing.

The idea is to allow the failover using the last restored PVC state back to the recovered cluster.

New Hub-


busybox-workloads-15   cephfs-sub-busybox15-placement-1-drpc         7d9h   amagrawa-c2-29a    amagrawa-c1-29a   Failover       FailedOver     WaitForReadiness   2024-05-17T07:41:59Z                        False


oc get drpc -o yaml -n busybox-workloads-15
apiVersion: v1
items:
- apiVersion: ramendr.openshift.io/v1alpha1
  kind: DRPlacementControl
  metadata:
    annotations:
      drplacementcontrol.ramendr.openshift.io/app-namespace: busybox-workloads-15
      drplacementcontrol.ramendr.openshift.io/last-app-deployment-cluster: amagrawa-c2-29a
    creationTimestamp: "2024-05-16T07:51:31Z"
    finalizers:
    - drpc.ramendr.openshift.io/finalizer
    generation: 3
    labels:
      cluster.open-cluster-management.io/backup: ramen
      velero.io/backup-name: acm-resources-generic-schedule-20240516070015
      velero.io/restore-name: restore-acm-acm-resources-generic-schedule-20240516070015
    name: cephfs-sub-busybox15-placement-1-drpc
    namespace: busybox-workloads-15
    ownerReferences:
    - apiVersion: cluster.open-cluster-management.io/v1beta1
      blockOwnerDeletion: true
      controller: true
      kind: Placement
      name: cephfs-sub-busybox15-placement-1
      uid: 31b90e55-e8e3-42b4-8f0a-ca8a71daa7ab
    resourceVersion: "36276430"
    uid: e0bb1638-5fa6-4b45-8a5e-2bc688c38101
  spec:
    action: Failover
    drPolicyRef:
      apiVersion: ramendr.openshift.io/v1alpha1
      kind: DRPolicy
      name: my-drpolicy-5
    failoverCluster: amagrawa-c1-29a
    placementRef:
      apiVersion: cluster.open-cluster-management.io/v1beta1
      kind: Placement
      name: cephfs-sub-busybox15-placement-1
      namespace: busybox-workloads-15
    preferredCluster: amagrawa-c2-29a
    pvcSelector:
      matchLabels:
        appname: busybox_app3_cephfs
  status:
    actionStartTime: "2024-05-17T07:41:59Z"
    conditions:
    - lastTransitionTime: "2024-05-17T07:42:28Z"
      message: Completed
      observedGeneration: 3
      reason: FailedOver
      status: "True"
      type: Available
    - lastTransitionTime: "2024-05-17T07:41:59Z"
      message: Started failover to cluster "amagrawa-c1-29a"
      observedGeneration: 3
      reason: NotStarted
      status: "False"
      type: PeerReady
    lastUpdateTime: "2024-05-23T16:40:50Z"
    phase: FailedOver
    preferredDecision:
      clusterName: amagrawa-c1-29a
      clusterNamespace: amagrawa-c1-29a
    progression: WaitForReadiness
    resourceConditions:
      conditions:
      - lastTransitionTime: "2024-05-16T08:00:28Z"
        message: All VolSync PVCs are ready
        observedGeneration: 6
        reason: Ready
        status: "True"
        type: DataReady
      - lastTransitionTime: "2024-05-16T08:00:28Z"
        message: Not all VolSync PVCs are protected
        observedGeneration: 6
        reason: DataProtected
        status: "False"
        type: DataProtected
      - lastTransitionTime: "2024-05-16T08:00:16Z"
        message: Nothing to restore
        observedGeneration: 6
        reason: Restored
        status: "True"
        type: ClusterDataReady
      - lastTransitionTime: "2024-05-16T08:00:28Z"
        message: Not all VolSync PVCs are protected
        observedGeneration: 6
        reason: DataProtected
        status: "False"
        type: ClusterDataProtected
      resourceMeta:
        generation: 6
        kind: VolumeReplicationGroup
        name: cephfs-sub-busybox15-placement-1-drpc
        namespace: busybox-workloads-15
        protectedpvcs:
        - busybox-pvc-1
kind: List
metadata:
  resourceVersion: ""






Recovered cluster C1-

oc project busybox-workloads-15; oc get pvc,vr,vrg,pods -o wide
Now using project "busybox-workloads-15" on server "https://api.amagrawa-c2-29a.qe.rh-ocs.com:6443".
NAME                                              STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS                    VOLUMEATTRIBUTESCLASS   AGE    VOLUMEMODE
persistentvolumeclaim/busybox-pvc-1               Bound    pvc-a333d21e-3ab7-425d-8254-8fa62522dc3f   94Gi       RWX            ocs-storagecluster-cephfs       <unset>                 23d    Filesystem
persistentvolumeclaim/volsync-busybox-pvc-1-src   Bound    pvc-06823313-ed2d-49df-9773-55ef9a56f114   94Gi       ROX            ocs-storagecluster-cephfs-vrg   <unset>                 7d9h   Filesystem

NAME                                                                                DESIREDSTATE   CURRENTSTATE
volumereplicationgroup.ramendr.openshift.io/cephfs-sub-busybox15-placement-1-drpc   primary        Primary

NAME                                            READY   STATUS    RESTARTS   AGE    IP             NODE        NOMINATED NODE   READINESS GATES
pod/busybox-1-7f9b67dc95-6hjn2                  1/1     Running   0          7d9h   10.128.3.234   compute-2   <none>           <none>
pod/volsync-rsync-tls-src-busybox-pvc-1-676pm   0/1     Error     0          26m    10.128.2.70    compute-2   <none>           <none>
pod/volsync-rsync-tls-src-busybox-pvc-1-zxzl5   1/1     Running   0          4m7s   10.128.2.71    compute-2   <none>           <none>




oc describe vrg
Name:         cephfs-sub-busybox15-placement-1-drpc
Namespace:    busybox-workloads-15
Labels:       <none>
Annotations:  drplacementcontrol.ramendr.openshift.io/destination-cluster: amagrawa-c2-29a
              drplacementcontrol.ramendr.openshift.io/do-not-delete-pvc:
              drplacementcontrol.ramendr.openshift.io/drpc-uid: e0bb1638-5fa6-4b45-8a5e-2bc688c38101
API Version:  ramendr.openshift.io/v1alpha1
Kind:         VolumeReplicationGroup
Metadata:
  Creation Timestamp:  2024-04-30T13:36:03Z
  Finalizers:
    volumereplicationgroups.ramendr.openshift.io/vrg-protection
  Generation:  6
  Owner References:
    API Version:     work.open-cluster-management.io/v1
    Kind:            AppliedManifestWork
    Name:            661184cbe6aabc283e2f4acb234afb291390b8b4b3dd10af342eca0c4e7e3f41-cephfs-sub-busybox15-placement-1-drpc-busybox-workloads-15-vrg-mw
    UID:             79905b6c-78f9-414c-abc9-a6506a5cf852
  Resource Version:  47644387
  UID:               61c8fe31-6d15-4b42-876e-2f5d9f8d55af
Spec:
  Action:  Failover
  Async:
    Replication Class Selector:
    Scheduling Interval:  5m
    Volume Snapshot Class Selector:
  Pvc Selector:
    Match Labels:
      Appname:        busybox_app3_cephfs
  Replication State:  primary
  s3Profiles:
    s3profile-amagrawa-c1-29a-ocs-storagecluster
    s3profile-amagrawa-c2-29a-ocs-storagecluster
  Vol Sync:
Status:
  Conditions:
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               All VolSync PVCs are ready
    Observed Generation:   6
    Reason:                Ready
    Status:                True
    Type:                  DataReady
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               Not all VolSync PVCs are protected
    Observed Generation:   6
    Reason:                DataProtected
    Status:                False
    Type:                  DataProtected
    Last Transition Time:  2024-05-16T08:00:16Z
    Message:               Nothing to restore
    Observed Generation:   6
    Reason:                Restored
    Status:                True
    Type:                  ClusterDataReady
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               Not all VolSync PVCs are protected
    Observed Generation:   6
    Reason:                DataProtected
    Status:                False
    Type:                  ClusterDataProtected
  Kube Object Protection:
  Last Update Time:     2024-05-23T16:40:25Z
  Observed Generation:  6
  Protected PV Cs:
    Access Modes:
      ReadWriteMany
    Annotations:
      apps.open-cluster-management.io/hosting-subscription:  busybox-workloads-15/cephfs-sub-busybox15-subscription-1
      apps.open-cluster-management.io/reconcile-option:      merge
    Conditions:
      Last Transition Time:  2024-05-16T08:00:16Z
      Message:               Ready
      Observed Generation:   6
      Reason:                SourceInitialized
      Status:                True
      Type:                  ReplicationSourceSetup
      Last Transition Time:  2024-05-16T07:59:24Z
      Message:               PVC restored
      Observed Generation:   5
      Reason:                Restored
      Status:                True
      Type:                  PVsRestored
    Labels:
      App:                                             cephfs-sub-busybox15
      app.kubernetes.io/part-of:                       cephfs-sub-busybox15
      Appname:                                         busybox_app3_cephfs
      apps.open-cluster-management.io/reconcile-rate:  medium
      velero.io/backup-name:                           acm-resources-schedule-20240516070016
      velero.io/restore-name:                          restore-acm-acm-resources-schedule-20240516070016
    Name:                                              busybox-pvc-1
    Namespace:                                         busybox-workloads-15
    Protected By Vol Sync:                             true
    Replication ID:
      Id:
    Resources:
      Requests:
        Storage:         94Gi
    Storage Class Name:  ocs-storagecluster-cephfs
    Storage ID:
      Id:
  State:   Primary
Events:
  Type    Reason                    Age                   From                               Message
  ----    ------                    ----                  ----                               -------
  Normal  PrimaryVRGProcessSuccess  62m (x42 over 3h21m)  controller_VolumeReplicationGroup  Primary Success
  Normal  PrimaryVRGProcessSuccess  20m (x5 over 62m)     controller_VolumeReplicationGroup  Primary Success



C1 still has replication source for the failedover workload but not replication destination

oc get replicationsources.volsync.backube -A
NAMESPACE              NAME            SOURCE          LAST SYNC   DURATION   NEXT SYNC
busybox-workloads-15   busybox-pvc-1   busybox-pvc-1




Surviving cluster C2-


oc project busybox-workloads-15; oc get pvc,vr,vrg,pods -o wide
Already on project "busybox-workloads-15" on server "https://api.amagrawa-c2-29a.qe.rh-ocs.com:6443".
NAME                                              STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS                    VOLUMEATTRIBUTESCLASS   AGE    VOLUMEMODE
persistentvolumeclaim/busybox-pvc-1               Bound    pvc-a333d21e-3ab7-425d-8254-8fa62522dc3f   94Gi       RWX            ocs-storagecluster-cephfs       <unset>                 23d    Filesystem
persistentvolumeclaim/volsync-busybox-pvc-1-src   Bound    pvc-06823313-ed2d-49df-9773-55ef9a56f114   94Gi       ROX            ocs-storagecluster-cephfs-vrg   <unset>                 7d9h   Filesystem

NAME                                                                                DESIREDSTATE   CURRENTSTATE
volumereplicationgroup.ramendr.openshift.io/cephfs-sub-busybox15-placement-1-drpc   primary        Primary

NAME                                            READY   STATUS    RESTARTS   AGE     IP             NODE        NOMINATED NODE   READINESS GATES
pod/busybox-1-7f9b67dc95-6hjn2                  1/1     Running   0          7d9h    10.128.3.234   compute-2   <none>           <none>
pod/volsync-rsync-tls-src-busybox-pvc-1-676pm   0/1     Error     0          28m     10.128.2.70    compute-2   <none>           <none>
pod/volsync-rsync-tls-src-busybox-pvc-1-zxzl5   1/1     Running   0          5m19s   10.128.2.71    compute-2   <none>           <none>


oc describe vrg
Name:         cephfs-sub-busybox15-placement-1-drpc
Namespace:    busybox-workloads-15
Labels:       <none>
Annotations:  drplacementcontrol.ramendr.openshift.io/destination-cluster: amagrawa-c2-29a
              drplacementcontrol.ramendr.openshift.io/do-not-delete-pvc:
              drplacementcontrol.ramendr.openshift.io/drpc-uid: e0bb1638-5fa6-4b45-8a5e-2bc688c38101
API Version:  ramendr.openshift.io/v1alpha1
Kind:         VolumeReplicationGroup
Metadata:
  Creation Timestamp:  2024-04-30T13:36:03Z
  Finalizers:
    volumereplicationgroups.ramendr.openshift.io/vrg-protection
  Generation:  6
  Owner References:
    API Version:     work.open-cluster-management.io/v1
    Kind:            AppliedManifestWork
    Name:            661184cbe6aabc283e2f4acb234afb291390b8b4b3dd10af342eca0c4e7e3f41-cephfs-sub-busybox15-placement-1-drpc-busybox-workloads-15-vrg-mw
    UID:             79905b6c-78f9-414c-abc9-a6506a5cf852
  Resource Version:  47644387
  UID:               61c8fe31-6d15-4b42-876e-2f5d9f8d55af
Spec:
  Action:  Failover
  Async:
    Replication Class Selector:
    Scheduling Interval:  5m
    Volume Snapshot Class Selector:
  Pvc Selector:
    Match Labels:
      Appname:        busybox_app3_cephfs
  Replication State:  primary
  s3Profiles:
    s3profile-amagrawa-c1-29a-ocs-storagecluster
    s3profile-amagrawa-c2-29a-ocs-storagecluster
  Vol Sync:
Status:
  Conditions:
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               All VolSync PVCs are ready
    Observed Generation:   6
    Reason:                Ready
    Status:                True
    Type:                  DataReady
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               Not all VolSync PVCs are protected
    Observed Generation:   6
    Reason:                DataProtected
    Status:                False
    Type:                  DataProtected
    Last Transition Time:  2024-05-16T08:00:16Z
    Message:               Nothing to restore
    Observed Generation:   6
    Reason:                Restored
    Status:                True
    Type:                  ClusterDataReady
    Last Transition Time:  2024-05-16T08:00:28Z
    Message:               Not all VolSync PVCs are protected
    Observed Generation:   6
    Reason:                DataProtected
    Status:                False
    Type:                  ClusterDataProtected
  Kube Object Protection:
  Last Update Time:     2024-05-23T16:40:25Z
  Observed Generation:  6
  Protected PV Cs:
    Access Modes:
      ReadWriteMany
    Annotations:
      apps.open-cluster-management.io/hosting-subscription:  busybox-workloads-15/cephfs-sub-busybox15-subscription-1
      apps.open-cluster-management.io/reconcile-option:      merge
    Conditions:
      Last Transition Time:  2024-05-16T08:00:16Z
      Message:               Ready
      Observed Generation:   6
      Reason:                SourceInitialized
      Status:                True
      Type:                  ReplicationSourceSetup
      Last Transition Time:  2024-05-16T07:59:24Z
      Message:               PVC restored
      Observed Generation:   5
      Reason:                Restored
      Status:                True
      Type:                  PVsRestored
    Labels:
      App:                                             cephfs-sub-busybox15
      app.kubernetes.io/part-of:                       cephfs-sub-busybox15
      Appname:                                         busybox_app3_cephfs
      apps.open-cluster-management.io/reconcile-rate:  medium
      velero.io/backup-name:                           acm-resources-schedule-20240516070016
      velero.io/restore-name:                          restore-acm-acm-resources-schedule-20240516070016
    Name:                                              busybox-pvc-1
    Namespace:                                         busybox-workloads-15
    Protected By Vol Sync:                             true
    Replication ID:
      Id:
    Resources:
      Requests:
        Storage:         94Gi
    Storage Class Name:  ocs-storagecluster-cephfs
    Storage ID:
      Id:
  State:   Primary
Events:
  Type    Reason                    Age                   From                               Message
  ----    ------                    ----                  ----                               -------
  Normal  PrimaryVRGProcessSuccess  63m (x42 over 3h22m)  controller_VolumeReplicationGroup  Primary Success
  Normal  PrimaryVRGProcessSuccess  20m (x5 over 62m)     controller_VolumeReplicationGroup  Primary Success


C2 has replication source too but that's expected (as failover is successful and workload is running on this cluster)

oc get replicationsources.volsync.backube -A
NAMESPACE              NAME            SOURCE          LAST SYNC   DURATION   NEXT SYNC
busybox-workloads-15   busybox-pvc-1   busybox-pvc-1


Expected results: Failover should complete with the last restored PVC state when replication is missing.


Additional info:
Comment 9 Sunil Kumar Acharya 2024-09-17 10:05:02 UTC 
Moving the non-blocker BZs out of ODF-4.17.0 as part of Development Freeze.
Note You need to log in before you can comment on or make changes to this bug.