There was a recent hotfix that Symantec release... it was mainly released to fix a bug in 12.5 where backing up Exchange 2 Tape, would results in on 32 cataloged mailboxes for GRT when you try to restore. The hotfix seems to have fixed this VSS Writer error for us. We also added 4 extra gig of ram to each cluster node upping it to 8 gig, as well as have patched exchange to rollup 6. Lastly we have a scheduled task that runs every Mon, Wed, Fri, which restarts the backup agent service on both exchange nodes. This is a seperate issue that we learned about with 12.5 where after about a week of backups on exchange the agent would not be releasing it's memory load and the beservice.exe (not sure of exact process name) would be using over 1+ gig of memory. This was obviously causing performance issues on the exchange side.
Hope this helps anyone that sees this!
