ESXi UNMAP not working on Replicated EqualLogic Volume

What to do when the VMware vSphere ESXi UNMAP command doesn’t release space on some or all volumes on a Dell EqualLogic SAN array running v8 firmware (may apply to other versions too).

Symptoms

  • Using the following command in an SSH session to a 6.0u2 host (again, will apply to other versions):
1esxcli storage vmfs unmap –l MYVOLUMENAME
  • The volumes are VMFS5 (and always have been- they haven’t been upgraded from VMFS3).
  • Replication is enabled for the volumes that won’t rethin.

Cause

UNMAP doesn’t work on the EqualLogic when Replication is enabled. It doesn’t return an error to the SSH session, and the temporary rethinning file is still created, but the disk is not thinned.

Solution

Disable replication on the volume, re-thin the volume using the UNMAP command, then re-configure replication. Unfortunately this means the entire volume must be re-copied to the replication partner and this may impact bandwidth usage and replication schedules on larger volumes.