1

A DFS-R is in place with 2 servers. ServerA in one location, ServerB on an another

Since yesterday morning we're in a situation where we don't know how to get rid of this.

The situation :

We receive a call where peoples don't have access anymore on a specific DFS namespace. Checking around, we found, on ServerA, that someone have first remove the share on this folder and 2 no more ACL are there.

First step, on the DFS namespace we disabled the Folder target on ServerA. To permit users accessing their files.

The replication looks to be very vey slow. drfsdiag backlog give 101000+ and 78000+ files on both side of the replication. Probably due to ACL mismatch I guess.

What is the best way to put back on track this situation ?

  • Stopping Replication of ServerA to ServerB ?
  • Delete the folder on ServerA ?
  • Recreate the replication with ServerB as primary source ?
  • Recreate ACL and share on ServerA ?
  • All of those steps in a specific order ?
  • Wait the backlogs is over and move after ?

Really new on DFS-R and we don't know how to proceed for that and all possible effect.

Any clue will be appreciate.

Patrick Pellegrino
  • 133
  • 3
  • 12

1 Answers1

0

If you changed ACLs on entire, large folders, you had to wait for the backlog to finish. For similar cases, please see here and here

If this is unacceptable, you can:

  • stop the replication
  • remove the affected folder (on a single side only!)
  • use robocopy to pre-seed the now empty destination
  • re-enable the replication

However, each of the above steps is not without downtime and, if something goes wrong, you can also worsen the situation. As always, take backup first.

Community
  • 1
shodanshok
  • 44,038
  • 6
  • 98
  • 162