|
Problem IC66367
|
Status: Closed |
STACK CORRUPTION WHEN CAPTURING A QUEUE THRESHOLD DEADLOCK USING THE LOCKING EVENT MONITOR |
| product: |
DB2 FOR LUW / DB2FORLUW / 970 - DB2 |
| Problem description: |
The DB2 server may terminate abnormally when capturing a queue
threshold deadlock if the locking event monitor is enabled.
The trap file produced should contain a call stack containing
the function sqmCollectQueueDeadlock.
sqmCollectDeadlockStackEntry
sqmCollectQueueDeadlock
sqlrwReportPotentialQueueBasedDeadlock
sqlpFindAndNotifyApp
sqlpReceiveGDDMessage
sqlpldl
sqloEDUEntry
Problem occurs only when using the locking event monitor and
thresholds. |
| Problem Summary: |
|
| Local Fix: |
NONE
Please note: An alternative method can be used to identify the
threshold and applications involved in the deadlock:
- Disable collection of deadlocks using the locking event
monitor by setting db cfg parameter to mon_deadlock to NONE
- Turn db cfg parameter mon_lck_msg_lvl to level 2
- Re-run application and monitor the db2diag.log or DB2
notification log
- A notification message will be generated when the queue
deadlock occurs. The message will contain information regarding
the threshold involved in the deadlock and the application
holding the lock
- If notification message is not enough to resolve the deadlock
and further diagnostic is required, then please upgrade to v9.7
Fixpack 2 and enable the lock event monitor to capture
deadlocks. |
| Solution |
|
| Workaround |
not known / see Local fix |
| Comment |
This APAR is a duplicate of IC66251 |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 16.02.2010
16.07.2010
16.07.2010 |
| Problem solved at the following versions (IBM BugInfos) |
|
| Problem solved according to the fixlist(s) of the following version(s) |
