Hi guys,
Could anybody give me an advice in the next situation please?
After power failure which affected Biztalk(2004) and SQL (2000) Servers multiple locks on BiztalkMsgBoxDB were found by DBA.
My findings:
1. Most of the locks belong to the object (Lock/Object in Enterprise Manager) BiztalkMsgBoxDB.dbo. BiztalkAppSendQ (host dedicated to sending operations through Covast EDI Accelerator although I don’t think it is relevant to the problem). The number of locks for this object is about 40000. Most of them have bts_DeQueueMessages_ BiztalkAppSendQ as a last operation.
2. The table BiztalkAppSendQ has approx 140000 rows.
3. SQL sysprocesses table has about 80 rows but no rows with blocked>0.
4. BizTalk server CPU utilization – 53%.
5. HAT – small reasonable number of dehydrated service instances and circa 500 ‘consumed’ message instances.
6. All Biztalk jobs are executing successfully.
It seems there is a problem with dequeueing messages from BiztalkAppSend. It is relatively long lasting problem so I can't rely on Biztalk itself. How to resolve it?
I am planning now a service window for SQL Server service restarting. What do you think if it is worth to do and it is enough?
Any ideas will be appreciated.