MQSeries.net :: View topic

klink · Posted: Fri Aug 01, 2003 5:36 am Post subject:

We are running WMQI 2.1 CSD03, MQ 5.3 fp3, AIX 5.1, DB2 7.2 fp9 client.

We have the configmgr running on win2k and the brokers running on AIX. The backend DB2 server is a remote server (to the AIX and NT Boxes) running on AIX as well.

I constantly (every day) have core dumps in every environment we are running. Currently we have Dev, Test, and Stage .... we go live in 3 weeks. Most of the time I get complete core dumps, but sometimes I don't. The DBX output is as follows. Obviously this is from an incomplete dump:

Several lines of
warning: Unable to access address 0x3000288c from core
warning: Unable to access address 0x3000288c from core

Followed by:

[using memory image in core]
warning: Unable to access address 0xf00cd304 from core
pthdb_session.c, 552: 1 PTHDB_CALLBACK (callback failed)
pthreaded.c, 1778: PTHDB_CALLBACK (callback failed)

IOT/Abort trap in pth_signal.pthread_kill [/usr/lib/libpthreads.a] at 0xd005a8f0
0xd005a8f0 (pthread_kill+0xa8) 80410014 lwz r2,0x14(r1)
(dbx)

I have my ulimits set to unlimited and I have plenty of file system space to hold the core. There is an associated wmqi abend file in the /var/mqsi/errors directory for each core. The contents of all are as follows:

abend record for pid 28984 tid 2057 time in seconds since 01/01/1970: 1059664705
File: /build/S210_P/src/CommonServices/Unix/ImbAbend.cpp
Line: 417
Function: signal received
---- Inserts ----
13
@(#) 1.28.1.4 CommonServices/Unix/ImbAbend.cpp, CommonServices, S210, S210-L2061
8 02/06/16 16:00:54 [6/18/02 14:41:19]
975066880
-----------------
----------------------------- Stack dump for current thread ( 2057)
(0xd0192798+0x000001ac) sqlcctcpsend__FP15sqlcc_comhandleP10sqlcc_cond [/home/db
2caed/sqllib/lib/libdtcp.a]
(0xd28a7e1c+0x0000044c) sqlccsend [/home/db2caed/sqllib/lib/db2.o.db2caef]
(0xd28ddb10+0x000000ec) sqle_db2ra_sendbuffer__FP17sqle_db2ra_commonP5sqlca [/ho me/db2caed/sqllib/lib/db2.o.db2caef]
(0xd2944e90+0x00002a20) sqle_db2ra_ar_sendrecv__FP10sqle_db2raP17sqle_ar_interfa
ceUli [/home/db2caed/sqllib/lib/db2.o.db2caef]
(0xd2940cbc+0x0000008c) sqle_db2ra_ar_prepexec__FP10sqle_db2raP17sqle_ar_interfa
ceUl [/home/db2caed/sqllib/lib/db2.o.db2caef]
(0xd2937ef0+0x00005e10) sqle_db2ra_ar_driver__FP10sqle_db2raP17sqle_ar_interface
[/home/db2caed/sqllib/lib/db2.o.db2caef]
(0xd28fbb04+0x000018d0) sqle_database_SQL__FP10sqle_db2raP10sqler_glob [/home/db

This is only part of the abend file. I will post more if necessary. We have DB2 setup with the Merant driver. I checked with Merant and there is no "updated" driver.

The core dump doesn't seem to interfere or otherwise affect the environment. No other errors in the system logs.

Weird thing is, I have the same issue with AIX 4.3.3, wmqi 2.1 csd04, mq 5.2 fp6, db2 7.2 fp7.

If anyone can help at all, it would be GREATLY appreciated!!!