Skip to main content

DRBD Recovery

This is documentation to bring back the old (Ubuntu 12.04) storage servers from a cold start to being able to connect with the XenServer cluster over NFS.

Current Configuration

Internal

  • Esau - primary/nfs
  • Jacob - secondary

External

  • Remus - primary/nfs
  • Romulus - secondary

The Steps

  • bring servers back from the dead, you can have them both up before starting anything
  • modprobe drbd - checks and enables the proper kernel module
  • drbd-overview - check drbd status
  • On Primary
    • drbdadm connect [i]nfs[1/2] - connect to the drbd shares
  • On Secondary
    • drbdadm – –discard-my-data connect [i]nfs[1/2] - connect to the drbd shares
  • On Primary
    • drbdadm primary [i]nfs[1/2] - set the primary server as the primary device within drbd
    • mount -o noatime /dev/drbd0 /srv/[i]nfs[1/2] - mount the drbd block device to the proper mount point
    • service nfs-kernel-service start - start the nfs service

You can now have the XenServer cluster go ahead and fix the NFS SR issues. Things should now be working.

Back to top