We are starting to experience a growing number of exclusive locks in tempdb. 7am things look fine, 8 am lots of locks. Some of our sql tools are complaining about queries taking longer to run than in prior runs. I found several implicit transactions where a "sleeping" connection maintains an open transaction which holds an X lock on some piece of tempdb. This can cause a "road jam" for other queries waiting to execute. While I can see evidence of this, I can't determine which query first caused this locking (to give the dev a hint what to look for in his code to refactor) Any ideas on that appreciated.
edm2