I have the following setup.
Server A transactionally replicates to Server B , using Server C as remote distributor.
Now the distribution agent runs on Server B, under distribution agents in replication monitor the distribution agent is the same job as the job on Server B defined as REPL-Dsitributor, correct ?
Now this morning at 4am under distribution agents on the remote distributor the job was marked as failed (error: no response in 10 minutes), but the "same job" on Server B, was not failed ?
The replication agent failure alert did not fire, neither did the notification failure alert on the job it self ?
What the %#$&#% happened ? This is a production environment so it is very worrying ?
Only according to the distribution agent on the remote distributor did the job fail, but not the job itself on the subscriber ?
Firstly what happened here ?
Secondly how can i make the agent keep retrying ? Once it failed it stayed failed, can i configure it to keep trying ?
Thanx
How can i make this agent keep retrying ?As for your case, distribution agent may not run on B but C (remote distributor) and replmonitor actually monitor agent job on C.
If your subscription is push, distribution agent will run under C. But if your subscription is pull, it will run under B.
About your another question, if agent job can keep retrying if it fails. That's depends, if it fails to access some meta data or subscription database, some access issue, it will retry five times and fail. Other problem cause the agent job fail, it just fail, until your solve the problem and restart it again. Otherwise, it won't run.
Hope it will help you some.
-Yunjing
No comments:
Post a Comment