|
Problem IC63225
|
Status: Closed |
DB2 INSTANCE CRASHES ON WINDOWS 64 WITH ACCESS ERROR FROM DB2AGENT IN SQLERISFMPVALIDFORUSE |
| product: |
DB2 FOR LUW / DB2FORLUW / 970 - DB2 |
| Problem description: |
The DB2 instance trapped with a SAP workload with access
error where 2 of the dbm configuration parameters are set to
KEEPFENCED=YES and FENCED_POOL value is significantly
smaller than the value of MAX_COORDAGENTS and
MAX_CONNECTIONS.
.
The db2 registry variable DB2_AGENT_CACHING_FMP is set to NO.
.
The workload contains concurrent connections from more than 10
database clients. There are calls to the SYSPROC.ADMIN_CMD
and SYSPROC.ENV_GET_SYS_INFO two routines they are both not
thread-safe.
.
The trapped db2agent has the following stack:
<StackTrace>
sqlerIsFmpValidForUse
sqlerFindPoolFmp
sqlerGetFmpFromPool
sqlerInvokeFencedRoutine
sqlrical
sqlriSectInvoke
sqlrr_process_execute_request
sqlrr_execute
sqljs_ddm_excsqlstt
sqljsParse
sqljsSqlam
sqljsDriveRequests
sqljsDrdaAsInnerDriver
sqljsDrdaAsDriver
sqeAgent::RunEDU
sqzEDUObj::EDUDriver
sqlzRunEDU
sqloSpawnEDU
</StackTrace> |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* DB2 instance with dbm cfg parameter KEEPFENCED=YES, and *
* registry variable DB2_AGENT_CACHING_FMP=NO and running *
* SYSPROC.ADMIN_CMD and other not thread-safe routines *
****************************************************************
* PROBLEM DESCRIPTION: *
* This trap happen on Windows because the IPC resource that is *
* associated with a fmp row is *
* cleaned up before the row is removed, and a db2agent is able *
* to find that row in the fmp *
* table. When the db2agent tries to determine if the fmp *
* exists or not using its the IPC resources, *
* it traped accessing a NULL pointer. *
* . *
* The registry setting of DB2_AGENT_CACHING_FMP=NO forces each *
* fmp worker to be returned *
* to the pool each time after it has finished executing a *
* routine for a db2agent. Thus this issue is *
* more noticeable under this setting. *
****************************************************************
* RECOMMENDATION: *
* Install V97 fixpack 1 *
**************************************************************** |
| Local Fix: |
Use the following step to make sure db2agents caches fmps:
db2stop force
db2set DB2_AGENT_CACHING_FMP=YES
db2start |
| available fix packs: |
DB2 Version 9.7 Fix Pack 1 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 2 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 3 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 3a for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 4 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 5 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 7 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 9a for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 6 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 8 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 9 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 10 for Linux, UNIX, and Windows
|
| Solution |
The fix allows DB2 to avoid such trap with added check on the
IPC resource pointer. |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 16.09.2009
15.01.2010
15.01.2010 |
| Problem solved at the following versions (IBM BugInfos) |
9.7.FP1
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 9.7.0.1
|
 |
