Hello,
I have replication (subscription with updateable subscribers) (pull) running on sql 2008 subscriber with 5 sql 2000 publishers. all seem well except the Queue agent stops at loss of network connectivity. It dosen't appear to restart until I manually restart it. It will run for hours until it stops or a shorter period. Usually, early in the morning before any server activity.
It is not all the Queue readers stopping at the same time, seems random.
I have looked for competing jobs or heavy server activity, I do not see any.
Per the error logs it is a network loss. (the network is a VPN and is know to have short down times.) (mostly seconds. not hours) (this is the reason I am doing queued transactions)
Is there a way to have the queue agent restart itself if down or to keep trying to start until the problem clears. if so what affect will it have on the publishers overhead.
Is there something else I am missing as I am fairly new to replication.
The other agents seem to just jug along doing their jobs and the trans actions are followed through.
Thanks for any help
Barry