November 5, 2009 at 12:35 pm
I've recently had a problem on a SQL2005 SP2 server returning the above message. It turned out the problem was related to the NICs and not SQL. Looking over the history, the NIC performance degraded over about 10 days and so did the general performance, but by the time the IOs were taking long enough to trigger the warning the users were having serious performance issues. This warning is controlled by Trace Flag 830, is there any way to change the default value of 15 seconds to get a warning earlier? I definitely don't want to turn the flag off.
Any ideas?
Leo Miller
SQL Services
Leo
Nothing in life is ever so complicated that with a little work it can't be made more complicated.
November 5, 2009 at 9:16 pm
Leo,
The threshold is currently fixed at 15 seconds, though this may become configurable in a future release or service pack.
There are many better ways of monitoring I/O performance than scanning the error log for messages such as this. A good place to start is the sys.dm_io_pending_io_requests dynamic management view.
Interested readers might also like to check http://blogs.msdn.com/chrissk/archive/2008/06/19/i-o-requests-taking-longer-than-15-seconds-to-complete-on-file.aspx before concluding that a stalled/stuck I/O problem exists. That article only applies to 2005 pre-SP3 (2008 was never affected).
Paul
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply