|
Problem IC72336
|
Status: Closed |
DB2/AIX PURESCALE: CLUSTER CACHING FACILITY CA-SERVER PROCESS TRAPS ON FIX PACK 2 |
| product: |
DB2 FOR LUW / DB2FORLUW / 980 - DB2 |
| Problem description: |
On DB2/AIX Version 9.8 Fix Pack 2 only (pureScale), the
ca-server process may trap on database activation due to
exhausting the CF worker thread stack area.
The mgmt_lwd_log.* file in the DIAGPATH will contain the
following entry:
PD_LEVEL_ERR: ca-mgmnt-lwd: Waitpid sees WTERMSIG: status is
(11)
errpt -a from the host where the CF is located which show that
the ca-server process abnormally terminated with a signal 11
(SIGSEGV).
LABEL: CORE_DUMP
...
SIGNAL NUMBER
11
...
PROGRAM NAME
ca-server
...
ADDITIONAL INFORMATION
_doprnt 7C
snprintf 18C
The Additional information contains a partial function call
stack traceback. Other stack tracebacks are possible. |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* Systems running version 9.8 Fix Pack 2 *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* Use documented workaround *
**************************************************************** |
| Local Fix: |
Increase the default CF worker thread stack size with the
following workaround:
db2set DB2_CA_ENV_STRING="AIXTHREAD_STK=2097152" |
| available fix packs: |
DB2 Version 9.8 Fix Pack 3 for Linux, UNIX, and Windows
DB2 Version 9.8 Fix Pack 4 for AIX and Linux
DB2 Version 9.8 Fix Pack 5 for AIX and Linux
|
| Solution |
Problem first fixed in DB2 UDB Version 9.8 Fix Pack 3 |
| Workaround |
Use documented workaround |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 01.11.2010
16.02.2011
16.02.2011 |
| Problem solved at the following versions (IBM BugInfos) |
9.8.FP3
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 9.8.0.3
|
 |
