hello mahesh (9/17/2012)
Jonathan Kehayias (9/17/2012)
hello mahesh (9/17/2012)
Still noluck. I tried searching for key words: 'dead' , 'victim' in xml and not found any.May be it got overlapped as the deadlock event occured couple of weeks back.
Thanks
If you restarted the instance since the deadlock occurred it won't be in the system_health event session since it is a memory resident target and not persisted across service failures/restarts.
The system was not restarted; but it is a very active OLTP system so I m guessing the event might got flushed out. Whats the amount of time or data that it hold in the queue?
The deadlocks are occurring in batch jobs and is a repeated phenomenon in the system, though for some reason it has not occurred in last two weeks. Do I need to set/check any settings in the server to capture deadlock?
Thanks
If it's not in the session and the server has been restarted, you are either:
A: on RTM version and not collecting the graph from a bug that was fixed and should apply the latest SP
B: having tons of other resource contention issues in the instance that you should look at
You can skip all of the Extended Events stuff if you use Event Notifications, which is what John sent you the code to do. I haven't looked at his solution, but if it is based on what I did at PASS I am guessing that he just added to the solution and it should allow you to persist any deadlock to a table for later review.
Jonathan Kehayias | Principal Consultant | MCM: SQL Server 2008
My Blog | Twitter | MVP Profile
Training | Consulting | Become a SQLskills Insider
Troubleshooting SQL Server: A Guide for Accidental DBAs[/url]