I'm getting following error from SQL Server 2014 Push Transactional Replication Distribution Agent:
Replication Distribution Subsystem: agent 'agent-name' scheduled for retry. TCP Provider: The semaphore timeout period has expired.
This it is caused by a transient network hiccup. However, after connection is back and other applications resume working the distribution agent still reports that it is retrying (unsuccessfully).
Agent 'agent-name' is retrying after an error. 6 retries attempted. See agent job history in the Jobs folder for more details
The problem is solved by removing replication and setting it up again with new snapshot transfer, but this is problematic on the production environment. Sometimes the problem was resolved by restarting the SQL Server Agent service. The TCP Chimney Offload is disabled. The distribution agent parameters are mostly default. The replication load is quite low mostly without undistributed commands in the queue. Any ideas what is the root cause and why the replication is impaired permanently and retries don't work?
