Lock:Deadlock Chain 17887493 1 0X01 4 myserver 10:16:46.210 Parallel query worker thread was involved in a deadlock 0 971497 102 - Resource type Exchange Lock:Deadlock Chain 17887491 1 0X01 4 myserver 10:16:46.210 Parallel query worker thread was involved in a deadlock 0 971497 102 - Resource type Exchange Lock:Deadlock Chain 17887489 1 0X01 4 myserver 10:16:46.210 Parallel query worker thread was involved in a deadlock 0 971497 102 - Resource type Exchange Lock:Deadlock Chain 17887487 1 0X01 4 myserver 10:16:46.210 Parallel query worker thread was involved in a deadlock 0 971497 102 - Resource type Exchange
#Key lock sql server deadlock update#
Both side are doing the simple update query in read commited, auto-commit mode. I captued some "Lock:Deadlock Chain" and "Lock:Deadlock",īut no "Deadlock Graph" was captured. I had tried to make processes update different position/slice. I had tried to rebuild the index a few times, with high and low fill factor.
Is there any way to prevent deadlock being happen? Or, how can reduce the number of locks it takes for one single row update without using cursor? (It does not make sense to me because I am updating one row at once) Then, I find that for each update query, it does take 100(and more) PAGE locks. I did some check and find that the deadlock happen on "PAGE" level, not at ROW/record level.
No trigger or table relation.etc on the table. UPDATE table SET WHERE PK is varchar(11), has a clustered index on it.
I have some funny deadlock caused by a stupid simple SQL UPDATE query, on a flat plain table, under default "READ COMMITED" transaction.
0 kommentar(er)
