I'm seeing the same problem every Sunday night at midnight on my (mostly unused) dev farm. At the stroke of midnight, something starts pummeling tempdb and fills up the log file within minutes, consuming all the disk space. I have to go in Moday morning and restart the SQL instance and it stops. The SAN admins are breathing down my neck to find out why this is happening as it's affecting our SAN performance for an unused dev system.
I've changed all the weekly SharePoint timer jobs so they do not start at midnight and still the same thing happens at the same time. The only thing at midnight in the event logs and SQL server logs is:
This instance of SQL Server has been using a process ID of 1220 since 10/8/2012 9:00:18 AM (local) 10/8/2012 2:00:18 PM (UTC). This is an informational message only; no user action is required.
Then at 12:31, I get error 17053 about disk space on the log file for tempdb. The log file is on a SAN mount, 37gb total, 94% free (right now after a restart). Should be plenty of room...
No scheduled tasks run on the server. My back ups are at 4am (nightly incremental and weekly full). Anything else I'm not thinking of?
Thanks.