Microsoft KB Archive/178138

{| The information in this article applies to:
 * width="100%"|
 * Microsoft SQL Server, version 6.5

Bug #: 17654 (NT: 6.5)

SYMPTOMS
After it cancels a striped table dump, Microsoft SQL Server may begin logging a recursive Msg 1203. This problem only occurs on a striped table dump, and the errors are logged repeatedly, until SQL Server is stopped.

The following sequence of events are recorded in the error log:

dmptable: User attention received Error : 1203, Severity: 20, State: 2 Caller of lock manager is incorrectly trying to unlock an unlocked object. spid=10 locktype=2 dbid=8 lockid=761105802.

The sequence of events repeats. This particular case (1203 error) will always have a locktype=2, a shared table lock; the lockid number identified in the error log will be the object id of the table which was being dumped.

CAUSE
A synchronization issue exists when multiple threads that service the dump try to release the same shared table lock. The error handling code incorrectly attempts to release the lock again, which causes the recurring error 1203.

Even when all of the above conditions are met, the problem does not occur frequently, because of the narrow time frame in which this problem can occur.

WORKAROUND
When you perform a table dump, specify only a single dump device or use the bulk copy program (BCP) utility to copy the table data.

STATUS
Microsoft has confirmed this to be a problem in Microsoft SQL Server version 6.5. We are researching this problem and will post new information here in the Microsoft Knowledge Base as it becomes available.
 * }

-

Last reviewed: December 16, 1997

© 1998 Microsoft Corporation. All rights reserved. Terms of Use.