FIX: SQL Terminates on Repeated Cursor Calls w/ ODS HandlersID: Q153917
|
A DB-Library application making repeated cursor calls encounters one or
more of the following errors:
SQL Server message 1101, state 1, severity 17: unable to allocate new
page for database 'tempdb'. There are no more pages available on valid
allocation pages.
space can be created by dropping objects, extending the database or
dumping the log with no_log.
DB-LIBRARY error: General SQL Server error: Check error messages from
SQL Server
SQL Server terminates when repetitive queries with sp_cursoropen/fetch are
called with ODS handlers installed.
The unhandled exception at the server causes the following lines in the
error log:
96/04/08 21:00:15.41 spid11
************************************************************************
*******
96/04/08 21:00:15.71 spid11 *
96/04/08 21:00:15.72 spid11 * STACK DUMP:
96/04/08 21:00:15.72 spid11 *
96/04/08 21:00:15.73 spid11 * Exception Address = 77F64F09
Stop any ODS applications, such as SQL Trace, that may be running at the server. This causes the ODS handlers to be de-installed and prevents the server from terminating.
Microsoft has confirmed this to be a problem in Microsoft SQL Server version 6.5. This problem has been corrected in U.S. Service Pack 2 for Microsoft SQL Server version 6.5. For more information, contact your primary support provider.
With any ODS handlers installed and DB-Library applications making repeated
dbopen/dbcursorfetchex calls in a loop, the server terminates. Windows NT
service Control Manager confirms that the service has been terminated, and
the above partial stack dump is seen in the error log.
Any ODS application, such as SQL Trace, can cause this problem. An
alternative to SQL Trace is to use server side tracing, using Trace Flags.
Additional query words: trace server av
Keywords : kbnetwork SSrvODS kbbug6.50 kbfix6.50.sp2
Version : 6.5
Platform : WINDOWS
Issue type :
Last Reviewed: March 31, 1999