Error:
Replication-replication Distribution Subsystem: agent (null) scheduled for retry. Query timeout expired.
Based on my research, this error appears when there is a lot of contention on the distribution database. I do see a distribution query running for over 10 hours. If I kill this transaction, won't this cause issues with the replication since this may be a transaction that is reading and copying data? Wouldn't killing it suspend that process and possibly cause corruption.
The publisher and subscribers are all online; these query timeout errors seem to be related to the distribution server and the distribution database contention. I'm trying to identify what's happening overall, since I have many publications that are getting timeouts and from digging into the servers, it only appears the distribution server is behind.