Network and Storage Protocols

Move vFiler1 to vFiler0


Hi community,

For one our customer we need to create a vFiler1 for integrate the Filer in a new AD domain.

vFiler0 is in other AD domain.

After that, we have to migrate CIFS from vfiler0 To vFiler1 (via SnapMirror or Robocopy) and migrate users from the old to the new AD domain. The vFiler1 will also host new ESX servers with few VM (SAN iSCSI).

Once completed, the customer wants that the vFiler1 replaces the vFiler0.

I know that when we'll destroy vFiler1, will not destroy any of the user data and all ressources will return to vFiler0.

But is it enough to replace vfiler1 by vfiler0 or will we have to reconfigure some things (ipspace, CIFS Setup for AD domain,....) ?

Is there any procédure to replace vFiler0 by vFiler1 (I have found one for the opposite (move vFiler0 to a vFiler1)  ?

Thanks for your answers





vfiler0 is a fixed vfiler (Physical controller).  To IPspace or not to IPspace really comes down to if you want a separate routing table...if a different network you probably want an IPspace... or if secure multi-tenancy.  If the same network and the same users it may not be needed...if the same default gateway to vfiler0 and vfiler1, you probably will keep the same default-ipspace, but if a different default gateway and IPspace will be easier to have a new routing table rather than have route add network to manage in the default-ipspace.

You can snapmirror/copy, but you can also assign the volume to vfiler1... or if you need to keep the copy in vfiler0, it would be more efficient to FlexClone the volume in vfiler0, then add the clone to vfiler1...then you can test access (ACLS/SIDS/Mappings) and when ready to cutover you can split the clone and keep access without having to wait for a mirror/copy to finish since the clone is near instant and available during the split.

You can move resources from vfiler0 to vfiler1, but vfiler0 still needs a root volume and IP address for physical controller managemement.