Time out occurred waiting for buffer latch type 4 - help with reading the error info
-
Hello all,
This issue is occurring in a multi-cpu 32GB server running SQL Server 2008R2 RTM on Windows Server Std SP2
The server hosts 8 client databases, each roughly 5GB, with 10-30 office users per day active in each database via web interface.
About 2x per day a message is written to the errorlog stating "A time-out occurred while waiting for buffer latch -- type 4, bp 0000000697FA8700, page 1:73346, stat 0xc0000b, database id: 12, allocation unit Id: 72057595133165568, task 0x0000000004AD3948 : 0, waittime 300, flags 0x1000001018, owning task 0x0000000000000000. Not continuing to wait."
I am confused by the error information. It states that the database id is 12, yet there is no object in that database with an id of 73346 - there is only the 1 datafile. Am I misreading the info?
No SQL dump is being written, the Server Event Log reports that this is Event ID = 846 and records almost the identical information and nothing more. The database checks out clean from multiple DBCC passes, index and table fragmentation is minimal, and no other process in any other database seems to be impacted. At least, no other error conditions are being thrown.
Online information regarding this error is rather sparse and I have not encountered this before. What else can I check? Has anyone encountered this issue and learned something new to share? I know that we need to upgrade SQL to SP1 but that's not possible until after June 1.
Any tips are most welcome, and thanks for your support.
Steve
-
sarmistead (3/28/2012)
Hello all,This issue is occurring in a multi-cpu 32GB server running SQL Server 2008R2 RTM on Windows Server Std SP2
The server hosts 8 client databases, each roughly 5GB, with 10-30 office users per day active in each database via web interface.
About 2x per day a message is written to the errorlog stating "A time-out occurred while waiting for buffer latch -- type 4, bp 0000000697FA8700, page 1:73346, stat 0xc0000b, database id: 12, allocation unit Id: 72057595133165568, task 0x0000000004AD3948 : 0, waittime 300, flags 0x1000001018, owning task 0x0000000000000000. Not continuing to wait."
I am confused by the error information. It states that the database id is 12, yet there is no object in that database with an id of 73346 - there is only the 1 datafile. Am I misreading the info?
No SQL dump is being written, the Server Event Log reports that this is Event ID = 846 and records almost the identical information and nothing more. The database checks out clean from multiple DBCC passes, index and table fragmentation is minimal, and no other process in any other database seems to be impacted. At least, no other error conditions are being thrown.
Online information regarding this error is rather sparse and I have not encountered this before. What else can I check? Has anyone encountered this issue and learned something new to share? I know that we need to upgrade SQL to SP1 but that's not possible until after June 1.
Any tips are most welcome, and thanks for your support.
Hi, I have never had that error before. As for the "73346" is not an object id, but a page number, located on the file 1. I found two links which I'm guessing you may have already checked, In case you have not checked I'm adding them here:
http://troubleshootingsql.com/category/troubleshooting-sql-issues/
For more info about how to use DBCC PAGE, I'm adding this link http://blogs.msdn.com/b/sqlserverstorageengine/archive/2006/06/10/625659.aspx.
I hope this will help you to resolve your issue, please let me know if I can help. Good luck!
-
Thanks for your help. I should have realized the error message was returning a page reference and not an object ref; your info helped me determine the table at issue.
I'll trace the connections that access the table in question and see if I can find anything there - perhaps someone is running a very large ad-hoc query while online users are active as well...
Again, thanks!
Steve
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply