DB2 FOR LUW / DB2FORLUW / 980 - DB2

When monitoring a process, the PID is searched for in the /proc 
filesystem. In a heavily loaded environment, this may lead to a 
rare timing hole where the PID would not be found under /proc. 
This leads to CF getting a false offline. This will result in 
the process being killed and failover initiated. 
 
The stack trace looks like: 
 
2011-08-26-03.28.29.166425+540 I584608A785 LEVEL: Event 
PID : 5439526 TID : 1 PROC : db2rocme 128 [db2inst1] 
INSTANCE: db2inst1 NODE : 128 
HOSTNAME: host1 
EDUID : 1 EDUNAME: db2rocme 128 [db2inst1] 
FUNCTION: DB2 UDB, oper system services, sqlossig, probe:10 
MESSAGE : Sending SIGKILL to the following process id 
DATA #1 : signed integer, 4 bytes 
7930068 
CALLSTCK: 
[0] 0x0900000001F376EC pdLog@glue678 + 0xD4 
[1] 0x0900000001F375E8 sqlossig + 0x64 
[2] 0x0900000001F37510 
sqlhaKillProcesses__FP18SQLHA_PROCESS_INFOUlbT2 + 0x3B4 
[3] 0x0000000100022BE8 rocmCAKillProcesses + 0x9C8 
[4] 0x0000000100011770 rocmCACleanup + 0x790 
[5] 0x0000000100003518 main + 0x1B38 
[6] 0x0000000100000320 __start + 0x98

**************************************************************** 
* USERS AFFECTED:                                              * 
* ALL                                                          * 
**************************************************************** 
* PROBLEM DESCRIPTION:                                         * 
* See Error Description                                        * 
**************************************************************** 
* RECOMMENDATION:                                              * 
* Upgrade to DB2 Version 9.8 Fix Pack 5 or higher              * 
****************************************************************

Not available

First Fixed in V9.8 FP 5

not known / see Local fix

Date  - problem reported    :
Date  - problem closed      :
Date  - last modified       :

13.09.2011
17.10.2012
17.10.2012

9.8.FP5