NSS volumes with snapshots enabled cannot be mounted after a migration or failover

After migrating a source workload that has NSS volumes, the volumes will have new names.  Custom scripts that use the old volume names do not function as expected.

After migrating a source workload that has NSS volumes, the volumes will have new names.  Custom scripts that use the old volume names do not function as expected.

The NSS file system maintains a single global per-machine namespace of NSS volume names. Snapshots of NSS volumes are represented in the system as NSS volumes; therefore, when an NSS snapshot is created NSS volumes in this snapshot are renamed by NSS.

After being transferred to the target by PlateSpin Migrate or Protect, NSS volumes keep their new volume names. PlateSpin Migrate/Protect will automatically modify the target machine's setup so that transferred volumes are still mounted to the same mount points. However, in the cases when a system uses custom scripts that operate with NSS volume names, these scripts may have to be modified manually or with the help of post-conversion scripts to reflect changed volume names.

The algorithm according to which new names are assigned to NSS volumes in a snapshot is as following:

1. Shrink the volume name to 12 characters and try to add the _SV prefix.
2. If the volume with this name already exists, shrink the original volume name to 9 symbols and try to add _SVXXX prefix where XXX is a sequence number.
3. Find the first sequence number not being used already.

Using snapshots to transfer volumes results in less data being transferred, faster transfers, and ability of transferring volumes that experience high write load. In the case of an NSS file system, however, using snapshots may cause the volume names to change according to the algorithm explained above.