|
Problem IT08799
|
Status: Closed |
HADR STANDBY MIGHT CRASH RARELY BY ACTIVATE DB AT STANDBY |
| product: |
DB2 FOR LUW / DB2FORLUW / A50 - DB2 |
| Problem description: |
HADR standby might crash rarely with sigsegv by activate
database after flashcopy at standby.
Symptom:
The associated trap file has a similar stack trace as below.
-----
0x0900000009029590 sqluProcessLoadStartLogRcd...
0x0900000009053DE4 sqlpRecDbRedo...
0x0900000007712898 sqlprProcDPSrec...
0x0900000007702DA4 sqlpPRecReadLog...
0x0900000006AF60FC sqlpParallelRecovery...
0x09000000073E96B8 sqleSubCoordProcessRequest...
0x09000000069D1140 RunEDU...
0x0900000007C57C2C EDUDriver...
0x0900000007C651C8 sqloEDUEntry...
-----
Root cause:
There is a problem where to start rolling forward log record(s)
of LOAD command at HADR standby.
This problem might occur rarely.
Condition of problem:
- LOAD command is executed at HADR primary
- Flash copy and "ACTIVATE DATABASE" at HADR standby |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* DB2 UDB Version 10.5 *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* Upgrade to Version 10.5 FixPack 7. *
**************************************************************** |
| Local Fix: |
"DEACTIVATE DATABASE" after doing a LOAD at HADR primary will
avoid the problem for crash recovery only, but it will not
help for database rollforward (DB RFWD). The only workaround
if you hit the problem in DB RFWD, is to restore the database
again FROM AN EARLIER BACKUP image, and then run RFWD again.
But note that the problem might still happen again with this
earlier backup image, however, since the conditions to hit
this are pretty rare, it would be very hard (but not
impossible) to have this problem with 2 different backups.
Another workaround is to take OFFLINE db backup, instead of
ONLINE backups. If you restore an offline backup and then run
rfwd, you will not hit this problem. |
| Solution |
Problem was first fixed in DB2 UDB Version 10.5 FixPack 7. |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 08.05.2015
19.01.2016
19.01.2016 |
| Problem solved at the following versions (IBM BugInfos) |
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 10.5.0.7
|
 |
