We have a number of dfsr setups that sync two file servers with each other in a full mesh replication. There are just under 100 of these setups, and we seem to be having a strange issue with a few of them. We have a scheduled task that essentially touches a file on each of the servers replicated folders. DFS replicates them to the other member server and we have a pretty good idea of how well the replication is acting. On 3 of the setups, the replication stops at midnight, files are time stamped 11:59 and remain so until the dfsr service is restarted in the morning. After it is restarted, dfsr appears to do an intial sync all over again. Where no, or very few files have actually been modified, the backlog goes up to about 278,000 files on one replicated folder.
The only event in the event logs that seems related is the defragment of the dfsr.db which happens at midnight.
Has anyone had similar issues? I'm afraid I will have to rebuild the dfsr.db from scratch. I don't see any 2212, or subsequent 2214 events, nor 301 & 302 events in the log. Is there anyway to reschedule the defrag of the dfsr.db to verify that this is what is triggering this?
Thanks