June 25, 2007 at 3:13 pm
Can anyone help with these EventLog errors? We are running mirroring with a witness on several databases on a production server. The problem starts with EventID 1479 (Connection to mirror instance had timed out after 10 seconds without a response). Then comes EventID 1442 (Database mirroring for database x is inactive). And finally EventID 1474 (The specified network name is no longer available for xxx (mirror instance)).
Weirdly, before the database fail-over, the EventLog reports "Starting up database x". The whole cycle only took some 30 seconds, but then repeated twice more.
I don't seen anything else in the System or Application Event Viewer that sheds any light. To me it doesn't seem to be a SQL problem , but a network connectivity problem.
There are no Jobs or Backups running during this period. I have custom Alerts which report the same sequence: Connection with mirror lost; Principal running exposed; No Quorum; Synchronizing Principal; Synchronized Principal with Witness. The event seems to happen at night; usually, but not always near midnight.
Could there be DNS or other network issues working here? Any ideas where I should look?
Thanks,
Elliott
April 28, 2010 at 3:42 pm
Hi, did you ever get a resolution to your issue? We are having the same exact problem.
thanks!
April 28, 2010 at 5:35 pm
No, I didn't get any suggestions on where to look. I still believe that it is a network connectivity problem, not a SQL Server problem. The problem is rare enough that we just work around it.
Elliott
May 18, 2012 at 6:43 am
I just had the exact same issue happen....i believe it to be a network issue, but unable to find exact cause. The auto fail over may not be the best idea if i continue to see these false alarms.
Anybody have any suggestions?
Thanks!
Scott
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply