Automation Suite on Linux installation guide

Repaving or updating the NFS server​

Guidelines​

1. It is good practice to back up the disk attached to the NFS server where you create the backup. You can find this information defined as the nfs.mountpoint key in the backup.json file.

2. To repave the NFS server, take the following steps:

   1. Mount the backup disk on the newly repaved machine on the same mount point as defined by the nfs.mountpoint key in the backup.json file.
   2. Update the NFS endpoint of the new NFS server defined as nfs.endpoint in backup.json.
   3. After repaving the NFS server, make sure you have followed the Setting up the external NFS server guidance.

3. Once the update to NFS server is complete, it is recommended to reboot the machine. Also make sure to enable the backup on cluster. For instructions, see Backing up the cluster.

Adding a new node to the cluster​

1. Add the FQDN or the IP address of the new node to the allowlist of the NFS server. For instructions, see Allowing NFS mount path access to all backup and restore nodes.
2. Enable the backup on the new node post-installation. For instructions, see Backing up the cluster.

Using a local disk backup​

1. On the backup cluster, take the following steps:

   1. Connect to the backup cluster server node.

      assignment

      ssh <backup-cluster-node>
      ssh <backup-cluster-node>
      

   2. Unmount the backup disk.

      assignment

      umount /backup
      umount /backup
      

   3. Detach the backup disk from the virtual machine. The following is an Azure example:

      assignment

      az vm disk detach \
        --resource-group "${resource_group_name}" \
        --vm-name "${node_name}" \
        --name "server0-CephBackupDisk"
      az vm disk detach \
        --resource-group "${resource_group_name}" \
        --vm-name "${node_name}" \
        --name "server0-CephBackupDisk"
      

2. On the restore cluster, take the following steps:

   1. Connect to the restore cluster server node.

      assignment

      ssh <restore-cluster-node>
      ssh <restore-cluster-node>
      

   2. Attach and mount the backup disk. The following is an Azure example:

      assignment

      az vm disk attach \
        --resource-group "${resource_group_name}" \
        --vm-name "${node_name}" \
        --name "server0-CephBackupDisk"
      az vm disk attach \
        --resource-group "${resource_group_name}" \
        --vm-name "${node_name}" \
        --name "server0-CephBackupDisk"
      

   3. If /backup is currently mounted, unmount it.

      assignment

      umount /backup
      umount /backup
      

   4. Remove any existing /backup entry from /etc/fstab.

      assignment

      sed -i.bak '/\/backup/d' /etc/fstab
      sed -i.bak '/\/backup/d' /etc/fstab
      

   5. Identify the newly attached disk.

      assignment

      lsblk
      lsblk
      

      Locate the unmounted disk name (for example, /dev/sdX).

   6. Rescan physical volumes.

      assignment

      pvscan
      pvscan
      

   7. Rename the volume group (VG) on the attached disk.

      assignment

      vg_uuid=$(pvs --noheadings -o vg_uuid /dev/sdX | awk '{print $1}')
      vgrename "$vg_uuid" backupvg_restored
      vg_uuid=$(pvs --noheadings -o vg_uuid /dev/sdX | awk '{print $1}')
      vgrename "$vg_uuid" backupvg_restored
      

   8. Activate the renamed volume group.

      assignment

      vgchange -ay backupvg_restored
      vgchange -ay backupvg_restored
      

   9. Ensure the mount directory exists.

      assignment

      mkdir -p /backup
      mkdir -p /backup
      

   10. Mount the logical volume.

       assignment

       mount /dev/backupvg_restored/backuplv /backup
       mount /dev/backupvg_restored/backuplv /backup
       

   11. Add a persistent entry in /etc/fstab.

       assignment

       echo '/dev/backupvg_restored/backuplv /backup xfs defaults 0 0' >> /etc/fstab
       echo '/dev/backupvg_restored/backuplv /backup xfs defaults 0 0' >> /etc/fstab
       

1. Verify that the/backup mount point is active.

   assignment

   df -h /backup
   df -h /backup
   

2. Confirm that backup data is accessible under /backup.

3. Proceed with restore operations as needed.