Non yielding scheduler dump during SQLBackup backup

ChrisAVWood

Hi,

We are running on SQL2005 SP2 build 3282 and SQLBackup 5.4 and had a non yielding scheduler dump with a Stack Signature for the dump of 0x00000337 last night. We had the same error 2 months ago when we were still running SQLBackup 5.3. I have raised Microsoft Connect issue 402603 but would like to see if this could be an SQLBackup problem as backup jobs with verify were running as scheduled jobs from SQL Agent when the condition occured both times.

Thanks

Chris

petey

A couple of customers had encoutered this issue previously. There are 2 knowledge base documents that seem relevant to the issue:

http://support.microsoft.com/kb/917971
http://support.microsoft.com/kb/810885

If your issue is due to a large msdb database, the resolution is then to reduce the size of the msdb database after clearing old data out of it. If your issue is due to a fast disk subsystem, one resolution that works for SQL Server 2005 is to reduce the number of threads that SQL Backup is using to back up the database. If you are backing up to multiple files, reduce the number of files. If you are backing up to a single file using multiple threads, reduce the number of threads used.

ChrisAVWood

It could be because of the size of msdb. I hadn't realized that it was as big as it was because it has DTA data in it.

Thanks

Chris

ChrisAVWood

Petey,

I had the same problem again this morning. It happens every 2 months. I had removed some DTA data from msdb to make its contents smaller but I had not shrunk msdb. The database is over 1.2 Gb but holds 200Mb of data. Can this problem occur even if the data content of msdb is nowhere near the actual size of msdb? Its also funny that it only happens once every 2 months and I can nothing extra happening on the day it occurs.

Thanks

Chris