| Author |
Message
|
| WBI_User21 |
 Posted: Tue Oct 28, 2008 1:18 am Post subject: AMQ6119 Error |
 |
|
Voyager
Joined: 12 Jun 2007
Posts: 98
|
Hi MQ Experts,
My Environment consists of Weblogic connecting to MQ 6.0 through JMS Bridges.As part of load testing , I configured 40+ concurrent clients and tried connecting to one of the QMGRs installed on Windows server 2003.
When I do this ,I am getting following FFST error
| Quote: |
+-----------------------------------------------------------------------------+
| |
| WebSphere MQ First Failure Symptom Report |
| ========================================= |
| |
| Date/Time :- Thu October 09 13:56:43 GMT Daylight Time 2008 |
| Host Name :- MYHOST(Windows Server 2003, Build 3790: Service |
| Pack 1) |
| PIDS :- 5724H7200 |
| LVLS :- 6.0.2.1 |
| Product Long Name :- WebSphere MQ for Windows |
| Vendor :- IBM |
| Probe Id :- XY129005 |
| Application Name :- MQM |
| Component :- xstConnectExtent |
| SCCS Info :- lib/cs/pc/winnt/amqxsten.c, 1.45.1.5 |
| Line Number :- 1013 |
| Build Date :- Mar 26 2007 |
| CMVC level :- p600-201-070323 |
| Build Type :- IKAP - (Production) |
| UserID :- SYSTEM |
| Process Name :- D:\bea\wls\920\WEBLOG~1\server\bin\beasvc.exe |
| Addressing mode :- 32-bit |
| Process :- 00055588 |
| Thread :- 00000001 |
| QueueManager :- QMGR |
| ConnId(1) IPCC :- 223 |
| Major Errorcode :- xecF_E_UNEXPECTED_SYSTEM_RC |
| Minor Errorcode :- OK |
| Probe Type :- MSGAMQ6119 |
| Probe Severity :- 2 |
| Probe Description :- AMQ6119: An internal WebSphere MQ error has occurred |
| (Failed to attach shared memory map: Handle(00000AD4) [rc=536895769 |
| errorCode=8]) |
| FDCSequenceNumber :- 0 |
| Arith1 :- 536895769 20006119 |
| Arith2 :- 8 8 |
| Comment1 :- Failed to attach shared memory map: Handle(00000AD4) |
| [rc=536895 |
| Comment2 :- Not enough storage is available to process this |
| command. |
| |
+----------------------------------------------------------------------------- |
If I do simple AMQSPUT , it works but when I try to do same with Weblogic following FFST gets generated.
I tried google with AMQ6119 error. This error found only with MQ 5 but never with MQ 6.
http://www-01.ibm.com/support/docview.wss?uid=swg21194875
As stated in above link, I checked the virtual memory of the box with help of admins. But VM is available.
Do I need to install any csd to fix this ? Did anyone faced this error in MQ 6.0 ? Please help .. |
|
| Back to top |
|
 |
| WBI_User21 |
| Posted: Wed Oct 29, 2008 12:06 am Post subject: |
|
|
Voyager
Joined: 12 Jun 2007
Posts: 98
|
I with help of admins monitored Memory and disk space utilization. We did not observed any problem in resources.
Can anyone help ? 
Thanks in advance.. |
|
| Back to top |
|
|
| Gaya3 |
| Posted: Wed Oct 29, 2008 12:13 am Post subject: |
|
|
Jedi
Joined: 12 Sep 2006
Posts: 2493
Location: Boston, US
|
I do remember Peter Potkay also experienced the same type of error before, don't know what action he has done to get rid off this issue.
He stated that its a constant problem of MQ V6.0.2.1,
It will be great if you can upgrade to MQ V6.0.2.4 and try
for more information: (just have a look)
http://www.mqseries.net/phpBB2/viewtopic.php?t=43399&highlight=&sid=f2531ac80d978bfd2bd24809bb7f4108
_________________
Regards
Gayathri
-----------------------------------------------
Do Something Before you Die |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Wed Oct 29, 2008 5:14 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
It was a memory leak in an old version of an exit. Once the exit was upgraded the problem went away. It was not a problem with MQ 6.0.2.1.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| WBI_User21 |
| Posted: Sun Nov 16, 2008 7:46 am Post subject: |
|
|
Voyager
Joined: 12 Jun 2007
Posts: 98
|
Hi Mr PeterPotkay,
Sorry for delayed reply .. .
What do you mean older version of exit ? Can you please throw some light on this ?
Thanks,
Ranga. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Sun Nov 16, 2008 8:42 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
We had an old version of an exit being used by the channel. If your channel definition does not specify an exit, this does not apply in your case.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Sun Nov 16, 2008 9:20 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
No more space in the 32 bit memory table. Check EXTSHM=ON.
Fixed in 6.0.2.3 (AIX at least) but I suspect you need to get to the right level on all platforms. Try upgrading to 6.0.2.5. A lot of the FDCs should disappear at this level. Enjoy 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| WBI_User21 |
| Posted: Wed Dec 03, 2008 8:25 am Post subject: |
|
|
Voyager
Joined: 12 Jun 2007
Posts: 98
|
This is what IBM suggested ..
The sequence of FDCs you received are normally indicative of a queue manager process being started by a logged on user instead of the MUSR_MQADMIN user.
If a process is started by a logged on user and it is the first process to connect and thus start a shared memory server, then that shared memory server will also start under the logged on user. As a result, when the user logs off, the shared memory server will be terminated, thus leading to the sequence of FDCs that you saw.
The most common occurrence of this error is when the queue manager itself is started from the command line under the logged on user by using strmqm. Always make sure that you start the qmgr through the Explorer, or with amqmdain from a prompt.
If you are unsure what id the MQ service is running under, please run the dcomcnfg check from the following link : http://www-1.ibm.com/support/docview.wss?rs=171&uid=swg21173090
Once you know the id, check task manager. All processes (with the exception of amqsvc and amqmtbrn) should be running under this id. If you find any exceptions, stop the service: net stop "IBM MQSeries"
Once the service ends, check task manager. If you find any MQ processes still running, end them. Then restart the service:
net start "IBM MQSeries"
Once the service restarts, check task manager again, and you should now find all MQ processes running under the MQ services id. |
|
| Back to top |
|
|
| Tibor |
| Posted: Wed Mar 24, 2010 8:09 am Post subject: |
|
|
Grand Master
Joined: 20 May 2001
Posts: 1033
Location: Hungary
|
I know it was an old topic for others but we were confronted with same problem two weeks ago. And yes, we found the APAR and the fixpak was installed but the problem happens every day since then. The only solution is the restarting of the application which caused this symptom after the memory allocation exceeds the 2 GB (platform: Windows 2003 Server x86, WebSphere MQ: v6.0.2.8, app: SiebelCRM v7.x)
But there was an enigmatic paragraph on the support side (http://www-01.ibm.com/support/docview.wss?uid=swg21315298):
APAR IC57176 was created to address this problem. The fix allows an option, IPCCManagedSets to be turned off for the Set management of the IPCC subpool used by the queue manager. This will make the initial extent to be mapped for the application will always be the minimum size.
I suppose there should be a configurable parameter somewhere in the registry but I didn't found any information about it.
Thanks in advance,
Tibor |
|
| Back to top |
|
|
|
|
