Chapter 10. Replacing a storage device

Procedure

1. Identify the OSD that needs to be replaced.

   # oc get -n openshift-storage pods -l app=rook-ceph-osd -o wide

   Example output:

   rook-ceph-osd-0-6d77d6c7c6-m8xj6    0/1    CrashLoopBackOff    0    24h   10.129.0.16   compute-2   <none>           <none>
   rook-ceph-osd-1-85d99fb95f-2svc7    1/1    Running    0    24h   10.128.2.24   compute-0   <none>           <none>
   rook-ceph-osd-2-6c66cdb977-jp542    1/1    Running    0    24h   10.131.2.32   compute-1   <none>           <none>

   In this example, rook-ceph-osd-0-6d77d6c7c6-m8xj6 needs to be replaced.

   Note

   If the OSD to be replaced is healthy, the status of the pod will be Running.

2. Scale down the OSD deployment for the OSD to be replaced

   # osd_id_to_remove=0
   # oc scale -n openshift-storage deployment rook-ceph-osd-${osd_id_to_remove} --replicas=0

   where, osd_id_to_remove is the integer in the pod name immediately after the rook-ceph-osd prefix. In this example, the deployment name is rook-ceph-osd-0.

   Example output:

   deployment.extensions/rook-ceph-osd-0 scaled

3. Verify that the rook-ceph-osd pod is terminated.

   # oc get -n openshift-storage pods -l ceph-osd-id=${osd_id_to_remove}

   Example output:

   No resources found.

   Note

   If the rook-ceph-osd pod is in terminating state, use the force option to delete the pod.

   # oc delete pod rook-ceph-osd-0-6d77d6c7c6-m8xj6 --force --grace-period=0

   Example output:

   warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
     pod "rook-ceph-osd-0-6d77d6c7c6-m8xj6" force deleted

4. Remove the old OSD from the cluster so that a new OSD can be added.

   # oc process -n openshift-storage ocs-osd-removal -p FAILED_OSD_ID=${osd_id_to_remove} | oc create -f -

   Warning

   This step results in OSD being completely removed from the cluster. Make sure that the correct value of osd_id_to_remove is provided.

5. Verify that the OSD is removed successfully by checking the status of the ocs-osd-removal pod. A status of Completed confirms that the OSD removal job succeeded.

   # oc get pod -l job-name=ocs-osd-removal-${osd_id_to_remove} -n openshift-storage

   Note

   If ocs-osd-removal fails and the pod is not in the expected Completed state, check the pod logs for further debugging. For example:

   # oc logs -l job-name=ocs-osd-removal-${osd_id_to_remove} -n openshift-storage --tail=-1

6. Delete the PVC resources associated with the OSD to be replaced.

   1. Identify the DeviceSet associated with the OSD to be replaced.

      # oc get -n openshift-storage -o yaml deployment rook-ceph-osd-${osd_id_to_remove} | grep ceph.rook.io/pvc

      Example output:

      ceph.rook.io/pvc: ocs-deviceset-0-0-nvs68
          ceph.rook.io/pvc: ocs-deviceset-0-0-nvs68

      In this example, the PVC name is ocs-deviceset-0-0-nvs68.

   2. Identify the PV associated with the PVC.

      # oc get -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix>

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in the previous step.

      Example output:

      NAME                      STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
      ocs-deviceset-0-0-nvs68   Bound    pvc-0e621d45-7d18-4d35-a282-9700c3cc8524   512Gi      RWO            thin           24h

      In this example, the PVC is ocs-deviceset-0-0-nvs68 that is identified in the previous step and associated PV is pvc-0e621d45-7d18-4d35-a282-9700c3cc8524.

   3. Identify the prepare-pod associated with the OSD to be replaced. Use the PVC name obtained in an earlier step.

      # oc describe -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix> | grep Mounted

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in an earlier step.

      Example output:

      Mounted By:    rook-ceph-osd-prepare-ocs-deviceset-0-0-nvs68-zblp7

   4. Delete the osd-prepare pod before removing the associated PVC.

      # oc delete -n openshift-storage pod rook-ceph-osd-prepare-ocs-deviceset-<x>-<y>-<pvc-suffix>-<pod-suffix>

      where, x, y, pvc-suffix, and pod-suffix are the values in the osd-prepare pod name identified in the previous step.

      Example output:

      pod "rook-ceph-osd-prepare-ocs-deviceset-0-0-nvs68-zblp7" deleted

   5. Delete the PVC associated with the device.

      # oc delete -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix>

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in an earlier step.

      Example output:

      persistentvolumeclaim "ocs-deviceset-0-0-nvs68" deleted

7. Create new OSD for new device.

   1. Delete the deployment for the OSD to be replaced.

      # oc delete -n openshift-storage deployment rook-ceph-osd-${osd_id_to_remove}

      Example output:

      deployment.extensions/rook-ceph-osd-0 deleted

   2. Verify that the PV for the device identified in an earlier step is deleted.

      # oc get -n openshift-storage pv pvc-0e621d45-7d18-4d35-a282-9700c3cc8524

      Example output:

      Error from server (NotFound): persistentvolumes "pvc-0e621d45-7d18-4d35-a282-9700c3cc8524" not found

      In this example, the PV name is pvc-0e621d45-7d18-4d35-a282-9700c3cc8524.

      • If the PV still exists, delete the PV associated with the device.

        # oc delete pv pvc-0e621d45-7d18-4d35-a282-9700c3cc8524

        Example output:

        persistentvolume "pvc-0e621d45-7d18-4d35-a282-9700c3cc8524" deleted

        In this example, the PV name is pvc-0e621d45-7d18-4d35-a282-9700c3cc8524.

   3. Deploy the new OSD by restarting the rook-ceph-operator to force operator reconciliation.

      1. Identify the name of the rook-ceph-operator.

         # oc get -n openshift-storage pod -l app=rook-ceph-operator

         Example output:

         NAME                                  READY   STATUS    RESTARTS   AGE
         rook-ceph-operator-6f74fb5bff-2d982   1/1     Running   0          1d20h

      2. Delete the rook-ceph-operator.

         # oc delete -n openshift-storage pod rook-ceph-operator-6f74fb5bff-2d982

         Example output:

         pod "rook-ceph-operator-6f74fb5bff-2d982" deleted

         In this example, the rook-ceph-operator pod name is rook-ceph-operator-6f74fb5bff-2d982.

      3. Verify that the rook-ceph-operator pod is restarted.

         # oc get -n openshift-storage pod -l app=rook-ceph-operator

         Example output:

         NAME                                  READY   STATUS    RESTARTS   AGE
         rook-ceph-operator-6f74fb5bff-7mvrq   1/1     Running   0          66s

         Creation of the new OSD may take several minutes after the operator restarts.

8. Delete the ocs-osd-removal job.

   # oc delete job ocs-osd-removal-${osd_id_to_remove}

   Example output:

   job.batch "ocs-osd-removal-0" deleted

Procedure

1. Identify the OSD that needs to be replaced and the OpenShift Container Platform node that has the OSD scheduled on it.

   # oc get -n openshift-storage pods -l app=rook-ceph-osd -o wide

   Example output:

   rook-ceph-osd-0-6d77d6c7c6-m8xj6    0/1    CrashLoopBackOff    0    24h   10.129.0.16   compute-2   <none>           <none>
   rook-ceph-osd-1-85d99fb95f-2svc7    1/1    Running    0    24h   10.128.2.24   compute-0   <none>           <none>
   rook-ceph-osd-2-6c66cdb977-jp542    1/1    Running    0    24h   10.130.0.18   compute-1   <none>           <none>

   In this example, rook-ceph-osd-0-6d77d6c7c6-m8xj6 needs to be replaced and compute-2 is the OCP node on which the OSD is scheduled.

   Note

   If the OSD to be replaced is healthy, the status of the pod will be Running.

2. Scale down the OSD deployment for the OSD to be replaced.

   # osd_id_to_remove=0
   # oc scale -n openshift-storage deployment rook-ceph-osd-${osd_id_to_remove} --replicas=0

   where osd_id_to_remove is the integer in the pod name immediately after the rook-ceph-osd prefix. In this example, the deployment name is rook-ceph-osd-0.

   Example output:

   deployment.extensions/rook-ceph-osd-0 scaled

3. Verify that the rook-ceph-osd pod is terminated.

   # oc get -n openshift-storage pods -l ceph-osd-id=${osd_id_to_remove}

   Example output:

   No resources found in openshift-storage namespace.

   Note

   If the rook-ceph-osd pod is in terminating state, use the force option to delete the pod.

   # oc delete pod rook-ceph-osd-0-6d77d6c7c6-m8xj6 --grace-period=0 --force

   Example output:

   warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
     pod "rook-ceph-osd-0-6d77d6c7c6-m8xj6" force deleted

4. Remove the old OSD from the cluster so that a new OSD can be added.

   1. Delete any old ocs-osd-removal jobs.

      # oc delete job ocs-osd-removal-${osd_id_to_remove}

      Example output:

      job.batch "ocs-osd-removal-0" deleted

   2. Remove the old OSD from the cluster

      # oc process -n openshift-storage ocs-osd-removal -p FAILED_OSD_ID=${osd_id_to_remove} | oc create -f -

      Warning

      This step results in OSD being completely removed from the cluster. Make sure that the correct value of osd_id_to_remove is provided.

5. Verify that the OSD is removed successfully by checking the status of the ocs-osd-removal pod. A status of Completed confirms that the OSD removal job succeeded.

   # oc get pod -l job-name=ocs-osd-removal-${osd_id_to_remove} -n openshift-storage

   Note

   If ocs-osd-removal fails and the pod is not in the expected Completed state, check the pod logs for further debugging. For example:

   # oc logs -l job-name=ocs-osd-removal-${osd_id_to_remove} -n openshift-storage --tail=-1

6. Delete the Persistent Volume Claim (PVC) resources associated with the OSD to be replaced.

   1. Identify the DeviceSet associated with the OSD to be replaced.

      # oc get -n openshift-storage -o yaml deployment rook-ceph-osd-${osd_id_to_remove} | grep ceph.rook.io/pvc

      Example output:

      ceph.rook.io/pvc: ocs-deviceset-0-0-nvs68
          ceph.rook.io/pvc: ocs-deviceset-0-0-nvs68

      In this example, the PVC name is ocs-deviceset-0-0-nvs68.

   2. Identify the PV associated with the PVC.

      # oc get -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix>

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in an earlier step.

      Example output:

      NAME                      STATUS        VOLUME        CAPACITY   ACCESS MODES   STORAGECLASS   AGE
      ocs-deviceset-0-0-nvs68   Bound   local-pv-d9c5cbd6   1490Gi     RWO            localblock     24h

      In this example, the associated PV is local-pv-d9c5cbd6.

   3. Identify the name of the device to be replaced.

      # oc get pv local-pv-<pv-suffix> -o yaml | grep path

      where, pv-suffix is the value in the PV name identified in an earlier step.

      Example output:

      path: /mnt/local-storage/localblock/nvme0n1

      In this example, the device name is nvme0n1.

   4. Identify the prepare-pod associated with the OSD to be replaced.

      # oc describe -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix> | grep Mounted

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in an earlier step.

      Example output:

      Mounted By:    rook-ceph-osd-prepare-ocs-deviceset-0-0-nvs68-zblp7

      In this example the prepare-pod name is rook-ceph-osd-prepare-ocs-deviceset-0-0-nvs68-zblp7.

   5. Delete the osd-prepare pod before removing the associated PVC.

      # oc delete -n openshift-storage pod rook-ceph-osd-prepare-ocs-deviceset-<x>-<y>-<pvc-suffix>-<pod-suffix>

      where, x, y, pvc-suffix, and pod-suffix are the values in the osd-prepare pod name identified in an earlier step.

      Example output:

      pod "rook-ceph-osd-prepare-ocs-deviceset-0-0-nvs68-zblp7" deleted

   6. Delete the PVC associated with the OSD to be replaced.

      # oc delete -n openshift-storage pvc ocs-deviceset-<x>-<y>-<pvc-suffix>

      where, x, y, and pvc-suffix are the values in the DeviceSet identified in an earlier step.

      Example output:

      persistentvolumeclaim "ocs-deviceset-0-0-nvs68" deleted

7. Replace the old device and use the new device to create a new OpenShift Container Platform PV.

   1. Log in to OpenShift Container Platform node with the device to be replaced. In this example, the OpenShift Container Platform node is compute-2.

      # oc debug node/compute-2

      Example output:

      Starting pod/compute-2-debug ...
      To use host binaries, run `chroot /host`
      Pod IP: 10.70.56.66
      If you don't see a command prompt, try pressing enter.
      # chroot /host

   2. Record the /dev/disk/by-id/{id} that is to be replaced using the device name, nvme0n1, identified earlier.

      # ls -alh /mnt/local-storage/localblock

      Example output:

      total 0
      drwxr-xr-x. 2 root root 51 Aug 18 19:05 .
      drwxr-xr-x. 3 root root 24 Aug 18 19:05 ..
      lrwxrwxrwx. 1 root root 57 Aug 18 19:05 nvme0n1 -> /dev/disk/by-id/nvme-eui.01000000010000005cd2e4de2f0f5251

   3. Find the name of the LocalVolume CR, and remove or comment out the device /dev/disk/by-id/{id} that is to be replaced.

      # oc get -n local-storage localvolume
      NAME          AGE
      local-block   25h

      # oc edit -n local-storage localvolume local-block

      Example output:

      [...]
          storageClassDevices:
          - devicePaths:
            - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4895e0e5251
            - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4ea2f0f5251
        #   - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4de2f0f5251
            storageClassName: localblock
            volumeMode: Block
      [...]

      Make sure to save the changes after editing the CR.

8. Log in to OpenShift Container Platform node with the device to be replaced and remove the old symlink.

   # oc debug node/compute-2

   Example output:

   Starting pod/compute-2-debug ...
   To use host binaries, run `chroot /host`
   Pod IP: 10.70.56.66
   If you don't see a command prompt, try pressing enter.
   # chroot /host

   1. Identify the old symlink for the device name to be replaced. In this example, the device name is nvme0n1.

      # ls -alh /mnt/local-storage/localblock

      Example output:

      total 0
      drwxr-xr-x. 2 root root 51 Aug 18 19:05 .
      drwxr-xr-x. 3 root root 24 Aug 18 19:05 ..
      lrwxrwxrwx. 1 root root 57 Aug 18 19:05 nvme0n1 -> /dev/disk/by-id/nvme-eui.01000000010000005cd2e4de2f0f5251

   2. Remove the symlink.

      # rm /mnt/local-storage/localblock/nvme0n1

   3. Verify that the symlink is removed.

      # ls -alh /mnt/local-storage/localblock

      Example output:

      total 0
      drwxr-xr-x. 2 root root 17 Apr 10 00:56 .
      drwxr-xr-x. 3 root root 24 Apr  8 23:03 ..

      Important

      For new deployments of OpenShift Container Storage 4.5 or later, LVM is not in use, ceph-volume raw mode is in play instead. Therefore, additional validation is not needed and you can proceed to the next step.

      For OpenShift Container Storage 4.4, or if OpenShift Container Storage has been upgraded to version 4.5 from a prior version, then both /dev/mapper and /dev/ should be checked to see if there are orphans related to ceph before moving on. Use the results of vgdisplay to find these orphans. If there is anything in /dev/mapper or /dev/ceph-* with ceph in the name that is not from the list of VG Names, use dmsetup to remove it.

9. Delete the PV associated with the device to be replaced, which was identified in earlier steps. In this example, the PV name is local-pv-d9c5cbd6.

   # oc delete pv local-pv-d9c5cbd6

   Example output:

   persistentvolume "local-pv-d9c5cbd6" deleted

10. Replace the device with the new device.

11. Log back into the correct OpenShift Container Platform node and identify the device name for the new drive. The device name can be the same as the old device, but the by-id must change unless you are reseating the same device.

    # lsblk

    Example output:

    NAME                         MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
    sda                            8:0    0   120G  0 disk
    |-sda1                         8:1    0   384M  0 part /boot
    |-sda2                         8:2    0   127M  0 part /boot/efi
    |-sda3                         8:3    0     1M  0 part
    `-sda4                         8:4    0 119.5G  0 part
      `-coreos-luks-root-nocrypt 253:0    0 119.5G  0 dm   /sysroot
    nvme0n1                      259:0    0   1.5T  0 disk

    In this example, the new device name is nvme0n1.

    1. Identify the /dev/disk/by-id/{id} for the new device and record it.

       # ls -alh /dev/disk/by-id | grep nvme0n1

       Example output:

       lrwxrwxrwx. 1 root root 57 Aug 18 19:05 nvme0n1 -> /dev/disk/by-id/nvme-eui.01000000010000005cd2e4ce090e5251

12. After the new /dev/disk/by-id/{id} is available a new disk entry can be added to the LocalVolume CR.

    1. Find the name of the LocalVolume CR.

       # oc get -n local-storage localvolume
       NAME          AGE
       local-block   25h

    2. Edit LocalVolume CR and add the new /dev/disk/by-id/{id}. In this example the new device is /dev/disk/by-id/nvme-eui.01000000010000005cd2e4ce090e5251.

       # oc edit -n local-storage localvolume local-block

       Example output:

       [...]
           storageClassDevices:
           - devicePaths:
             - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4895e0e5251
             - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4ea2f0f5251
         #   - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4de2f0f5251
             - /dev/disk/by-id/nvme-eui.01000000010000005cd2e4ce090e5251
             storageClassName: localblock
             volumeMode: Block
       [...]

       Make sure to save the changes after editing the CR.

13. Verify that there is a new PV in Available state and of the correct size.

    # oc get pv | grep 1490Gi

    Example output:

    local-pv-3e8964d3                          1490Gi      RWO            Delete           Bound       openshift-storage/ocs-deviceset-2-0-79j94   localblock                             25h
    local-pv-414755e0                          1490Gi      RWO            Delete           Bound       openshift-storage/ocs-deviceset-1-0-959rp   localblock                             25h
    local-pv-b481410                           1490Gi      RWO            Delete           Available

14. Create new OSD for new device.

    1. Delete the deployment for the OSD to be replaced.

       # osd_id_to_remove=0
       # oc delete -n openshift-storage deployment rook-ceph-osd-${osd_id_to_remove}

       Example output:

       deployment.extensions/rook-ceph-osd-0 deleted

    2. Deploy the new OSD by restarting the rook-ceph-operator to force operator reconciliation.

       1. Identify the name of the rook-ceph-operator.

          # oc get -n openshift-storage pod -l app=rook-ceph-operator

          Example output:

          NAME                                  READY   STATUS    RESTARTS   AGE
          rook-ceph-operator-6f74fb5bff-2d982   1/1     Running   0          1d20h

       2. Delete the rook-ceph-operator.

          # oc delete -n openshift-storage pod rook-ceph-operator-6f74fb5bff-2d982

          Example output:

          pod "rook-ceph-operator-6f74fb5bff-2d982" deleted

          In this example, the rook-ceph-operator pod name is rook-ceph-operator-6f74fb5bff-2d982.

       3. Verify that the rook-ceph-operator pod is restarted.

          # oc get -n openshift-storage pod -l app=rook-ceph-operator

          Example output:

          NAME                                  READY   STATUS    RESTARTS   AGE
          rook-ceph-operator-6f74fb5bff-7mvrq   1/1     Running   0          66s

          Creation of the new OSD may take several minutes after the operator restarts.