Database Mirroring virtual logs backup RRS feed

  • Question

  • We're having some issues with our SQL mirroring setup, I make no claims at being a DBA, however the DBA's at my company have been unable to provide any real useful information. They only see that SQL just seems to stop sending logs to the mirror endpoint, causing a backup of logs files and message like:

    Database [xxx] has more than 1000 virtual log files which is excessive. Too many virtual log files can cause long startup and backup times. Consider shrinking the log and using a different growth increment to reduce the number of virtual log files.

    I don't see anything in event logs on the primary clustered servers, nor the single passive server to indicate any issue. There's really very little to explain the issues they're seeing, and I don't have much access to the SQL 2008 R2 server itself (just the Windows 2008 R2 SP1 hosting the SQL) to poke around. I have to believe there are either logs other than the Windows event logs or some additional logging that can be enabled to help point me in the right direction.

    Server resources don't seem to be a constraint, the primary servers have 4x 8-core CPU's and 256GB of RAM, which matches the mirror node. The two data centers are connected via a 10GbE MAN (Metro-Area-Network) which provides <5ms of latency so that should be more than sufficient.

    Please let me know where else I can start looking to try and point to the cause.


    Thursday, April 25, 2013 11:32 PM