| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Has anyone successfully implemented WMQI 2.1 in an HA env? |
« View previous topic :: View next topic » |
| Author |
Message
|
| claire |
 Posted: Fri Jun 21, 2002 4:22 pm Post subject: Has anyone successfully implemented WMQI 2.1 in an HA env? |
 |
|
Apprentice
Joined: 10 Apr 2002
Posts: 32
|
Hi,
On AIX 4.3.3, I use SupportPac IC61 to set up HACMP for WMQI V2.1.
After I complete the setup,the broker can be started normally at the primary node.
Then, I stop the broker at the primary node and try to start the broker at the secondary node,
the following error message occurs in /var/log/syslog.user:
Jun 21 21:22:00 fepbk WMQIv210[49408]: (BROKER1.service)[1]BIP2228E: Severe erro
r: /build/S210_P/src/CommonServices/Unix/ImbGlobalMutexHelper.cpp 422 semctl Abe
nd file: /var/wmqi/errors/BROKER1.service.49408.1.abend action: abort
And the content of /var/wmqi/errors/BROKER1.service.49408.1.abend is listed here:
abend record for pid 49408 tid 1 time Fri Jun 21 21:22:00 2002
File: /build/S210_P/src/CommonServices/Unix/ImbGlobalMutexHelper.cpp
Line: 422
Function: semctl
---- Inserts ----
13
30
doInitialisation
@(#) 1.10.1.1 CommonServices/Unix/ImbGlobalMutexHelper.cpp, porting, S210, S210-L011123 01/11/18 07:17:07 [11/23/01 16:57:48]
-----------------
----------------------------- Stack dump for current thread ( 1)
(malloc allowed)
(0xd158f758+0x00000204) dumpCurrentStacksToFile__8ImbAbendFPcPCv [/usr/lib/libCommonServices.a(libCommonServices.a.so)]
(0xd158ce20+0x00000adc) logSystemErrorInternal__8ImbAbendFPCcCiPCcRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertPCvCb [/usr/lib/libCommonServices.a(libCommonServices.a.so)]
(0xd1587f90+0x0000010c) logSystemErrorAndTerminateProcessInternal__8ImbAbendFPCcCiPCcRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertRCQ2_8ImbAbend6InsertPCvCb [/usr/lib/libCommonServices.a(libCommonServices.a.so)]
(0xd16d76d4+0x00000b90) doInitialisation__20ImbGlobalMutexHelperFC13ImbStringBaseXTwTQ2_3std11char_traitsXTw_SP18SP37_ [/usr/lib/libCommonServices.a(libCommonServices.a.so)]
(0xd16d7634+0x00000018) initialize__20ImbGlobalMutexHelperFC13ImbStringBaseXTwTQ2_3std11char_traitsXTw_SP18SP37_ [/usr/lib/libCommonServices.a(libCommonServices.a.so)]
(0x10003220+0x00000a94) main [bipservice]
(0x10000128+0x0000008c) __start [bipservice]
----------------------------------------------------------------------
------------------------------------------ Stack dumps for all threads
(malloc allowed)
=== Thread 1
No stack information available
----------------------------------------------------------------------
Please give me some advice to solve this problem.
Thanks a lot. |
|
| Back to top |
|
 |
| CodeCraft |
| Posted: Fri Jun 21, 2002 11:07 pm Post subject: |
|
|
Disciple
Joined: 05 Sep 2001
Posts: 195
|
I haven't done this, and I'd suggest that it's a topic of complexity beyond the board since there's so much you can potentially get wrong in the HA setup.
The fact that there is an abend in mutex related code suggests that something in node b is still competing for resources with something in node a. Are you sure everything in a was shutdown cleanly before attempting to start up b?
This is most likely one for IBM support! |
|
| Back to top |
|
|
| TorpedoSTu |
| Posted: Tue Jun 25, 2002 6:20 am Post subject: |
|
|
Acolyte
Joined: 14 Nov 2001
Posts: 73
|
When you say 'I stopped the broker on Node1' - do you mean you've used HA failover ? - I don't think you can just start the broker on node 2 as db2 (?) and the queue manager are both still running on node 1.
Are you using the HACMP_MQSISTART scripts ?
Regs
Stuart |
|
| Back to top |
|
|
| claire |
| Posted: Tue Jun 25, 2002 1:27 pm Post subject: Has anyone successfully implemented WMQI 2.1 in an HA env? |
|
|
Apprentice
Joined: 10 Apr 2002
Posts: 32
|
In fact,I stop broker(mqsistop BROKER1),queue manager(endmqm BROKERQM1) and db2 server(db2stop) on node 1. Then,I unmount the related file systems and varyoff the volume group on node 1.
Finally,I issue the following commands on node 2:
root:
varyonvg havg
mount /MQHA/BROKERQM1/data
mount /MQHA/BROKERQM1/log
mount /brkdb
mqm:
strmqm BROKERQM1
db2inst1:
db2start
mqsistart BROKER1
But the broker cannot start on node 2. |
|
| Back to top |
|
|
| TorpedoSTu |
| Posted: Wed Jun 26, 2002 6:03 am Post subject: |
|
|
Acolyte
Joined: 14 Nov 2001
Posts: 73
|
Did you install the binaries on both nodes ?
Is the directory structure /var/msqi within your file systems you mention ?
Are there any shared memory segments (etc) left (see the old mq ha support pack)
Regs
Stu
Also, has db2inst1 got the privileges to start the broker ? or is htis a type in your message ? |
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
