I’ve noticed a bit of a slowdown in replication. The two servers are clustered. When watching the replicator task run, I see it continue through multiple databases normally. After replication completes I check the log and notice that only a small handful of databases were logged. Since most of the databases, about 90%, are not showing up in replication events, does this mean they were not replicated completely? Cluster replication appears to go fast and when checking the cluster replication events I can see a lot more databases listed and a much, much quicker replication time. In the connection doc, there is no time limit set for replication, it’s to replicate all databases, replicate databases of low, med, and high priority, the schedule is enabled for all days repeating every 30 minutes, the ip is correct for the destination server and usage priority is normal. I’m not sure but the slow down in replication appears after upgrading to 8.5.2.
Subject: replica stubs
I also don’t see any replica stubs out there and the few errors that do show up in the replication events are being/have been taken care of. Any errors that did come up seemed very minimal (replication disabled for abc.nsf, a mail file had an attachment problem that caused replication to only complete partially). My main concern right now is why the logged replication events are missing a huge chunk of the databases that should be replicated/logged.