| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| MQ Application problems on Tandem |
« View previous topic :: View next topic » |
| Author |
Message
|
| AndyMQ |
 Posted: Tue Nov 15, 2005 6:38 am Post subject: MQ Application problems on Tandem |
 |
|
Apprentice
Joined: 22 Apr 2004
Posts: 33
Location: Scotland
|
I've sent this on to IBM, but just in case it rings any bells with anyone.
We run MQ 5.1, CSD03, C3EFIX1 on Tandem G06.25 OS level.
We have ported an app from an old G06.21 system running MQ 5.1, CSD02, and we get the following messages in EMS....
05-11-15 11:48:05 \T4.$HJP0 *IBM.MQSeries.2 000006 MQSeries
Error
Error Code : 8192
Function : xstConnectNSKSegment
Probe ID : 3
FFST File : $SW2.ZMQSSYS.FD55141
05-11-15 11:48:05 \T4.$HJP0 *IBM.MQSeries.2 000001 MQSeries FFST
from component COMP_XC
Error Code : 6118
Severity : error
Module Name : xstConnectNSKSegment
Probe ID : 3
Error Text :
Program Abended xstConnectNSKSegment
3 $SW2.ZMQSSYS.FD55141
05-11-15 11:48:05 \T4.$HJP0 *IBM.MQSeries.2 000006 MQSeries
Error
Error Code : 8192
Function : xxxInitialize
Probe ID : 2
FFST File : $SW2.ZMQSSYS.FD55141
05-11-15 11:48:05 \T4.$HJP0 *IBM.MQSeries.2 000001 MQSeries FFST
from component COMP_XC
Error Code : 6118
Severity : error
Module Name : xxxInitialize
Probe ID : 2
Error Text :
Program Abended xxxInitialize
2 $SW2.ZMQSSYS.FD55141
The FD file created has multiple occurences of the following 2 messages:-
| MQSeries First Failure Symptom Report |
| ===================================== |
| |
| Date/Time :- November 15 9:43:31 2005 |
| Host Name :- \T4 |
| PIDS :- 5724A39 |
| LVLS :- 510 |
| Product Long Name :- MQSeries for Compaq NonStop Kernel |
| Vendor :- IBM |
| Probe Id :- XC142003 |
| Application Name :- MQM |
| Component :- xstConnectNSKSegment |
| Build Date :- Oct 7 2002 |
| Exe File Name :- \T4.$SW3.HEXECUTE.ASZJIZE |
| UserID :- -29441 |
| Process File Name :- \T4.$HJI0:6233201 |
| Node number :- 44 |
| CPU :- 5 |
| PIN :- 764 |
| QueueManager :- SSECREST01QM |
| Major Errorcode :- xecF_E_UNEXPECTED_RC |
| Minor Errorcode :- Unknown(1) |
| Probe Type :- MSGAMQ6118 |
| Probe severity :- Severity 2: error |
| Probe Description :- AMQ6118: An internal MQSeries error has occurred (1) |
| Text :- Failed to allocate NSK Segment |
| |
| Arith1 :- 1 (0x1) |
| Arith2 :- 48 (0x30) |
| Comment1 :- $DBMQ.SSECRESD.CCSIDMEM |
| |
| |
+-----------------------------------------------------------------------------+
+-----------------------------------------------------------------------------+
| |
| MQSeries First Failure Symptom Report |
| ===================================== |
| |
| Date/Time :- November 15 9:43:31 2005 |
| Host Name :- \T4 |
| PIDS :- 5724A39 |
| LVLS :- 510 |
| Product Long Name :- MQSeries for Compaq NonStop Kernel |
| Vendor :- IBM |
| Probe Id :- XC328002 |
| Application Name :- MQM |
| Component :- xxxInitialize |
| Build Date :- Oct 7 2002 |
| Exe File Name :- \T4.$SW3.HEXECUTE.ASZJIZE |
| UserID :- -29441 |
| Process File Name :- \T4.$HJI0:6233201 |
| Node number :- 44 |
| CPU :- 5 |
| PIN :- 764 |
| QueueManager :- SSECREST01QM |
| Major Errorcode :- xecF_E_UNEXPECTED_RC |
| Minor Errorcode :- xecI_S_FAIL |
| Probe Type :- MSGAMQ6118 |
| Probe severity :- Severity 2: error |
| Probe Description :- AMQ6118: An internal MQSeries error has occurred (4000 |
| 6004) |
| Text :- Failed to connect to CCSID shared memset |
| |
| Arith1 :- 1073766404 (0x40006004) |
| |
From the first message, it would suggest we have a security violation (arith2 = 4 connecting to the CCSIDMEM file. I've checked Guardian, Safeguard, and OAM settings, and all SEEM tickety-boo.
FInd it strange that the FD file is being generated at MQ level, rather than the actual queue manager that the program is attempting to connect to, but params and defines also seem fine.
Any ideas anyone? |
|
| Back to top |
|
 |
| LuisFer |
| Posted: Tue Nov 15, 2005 9:12 am Post subject: Re: MQ Application problems on Tandem |
|
|
Partisan
Joined: 17 Aug 2002
Posts: 302
|
I'm obtained this FDC running a program from other Node.
The tacl Node is \DS4 , the program is on \DS5 vol.subvol the QMgr T43 is on \DS4 node
\DS4.$SPOOL1.MQMMGR 1> run \ds5.$system.zmqsexe.mqping T43 T43
+-----------------------------------------------------------------------------+
| |
| MQSeries First Failure Symptom Report |
| ===================================== |
| |
| Date/Time :- November 15 17:59:26 2005 |
| Host Name :- \DS5 |
| PIDS :- 5724A39 |
| LVLS :- 510 |
| Product Long Name :- MQSeries for Compaq NonStop Kernel |
| Vendor :- IBM |
| Probe Id :- XC142003 |
| Application Name :- MQM |
| Component :- xstConnectNSKSegment |
| Build Date :- Jul 14 2005 |
| Exe File Name :- \DS5.$SYSTEM.ZMQSEXE.MQPING |
| UserID :- -3585 |
| Process File Name :- \DS5.$:1:1182:6370300 |
| Node number :- 4 |
| CPU :- 1 |
| PIN :- 1182 |
| QueueManager :- T43 |
| Major Errorcode :- xecF_E_UNEXPECTED_RC |
| Minor Errorcode :- Unknown(1) |
| Probe Type :- MSGAMQ6118 |
| Probe severity :- Severity 2: error |
| Probe Description :- AMQ6118: An internal MQSeries error has occurred (1)
|
| Text :- Failed to allocate NSK Segment |
| |
| Arith1 :- 1 (0x1) |
| Arith2 :- 20 (0x14) |
| Comment1 :- $SPOOL1.T43D.CCSIDMEM |
| |
| |
+-----------------------------------------------------------------------------+
2005/11/15 17.59.26 AMQ6118: An internal MQSeries error has occurred (1)
+-----------------------------------------------------------------------------+
| |
| MQSeries First Failure Symptom Report |
| ===================================== |
| |
| Date/Time :- November 15 17:59:26 2005 |
| Host Name :- \DS5 |
| PIDS :- 5724A39 |
| LVLS :- 510 |
| Product Long Name :- MQSeries for Compaq NonStop Kernel |
| Vendor :- IBM |
| Probe Id :- XC328002 |
| Application Name :- MQM |
| Component :- xxxInitialize |
| Build Date :- Jul 14 2005 |
| Exe File Name :- \DS5.$SYSTEM.ZMQSEXE.MQPING |
| UserID :- -3585 |
| Process File Name :- \DS5.$:1:1182:6370300 |
| Node number :- 4 |
| CPU :- 1 |
| PIN :- 1182 |
| QueueManager :- T43 |
| Major Errorcode :- xecF_E_UNEXPECTED_RC |
| Minor Errorcode :- xecI_S_FAIL |
| Probe Type :- MSGAMQ6118 |
| Probe severity :- Severity 2: error |
| Probe Description :- AMQ6118: An internal MQSeries error has occurred (4000 |
| 6004)
|
| Text :- Failed to connect to CCSID shared memset |
| |
| Arith1 :- 1073766404 (0x40006004) |
| |
+-----------------------------------------------------------------------------+
The manual says:
Guardian segment IDs used by MQSeries
MQSeries allocates Guardian memory segments both in its own processes and in
the application program’s process. The Guardian Segment IDs used by MQSeries
for these segments are allocated from a range defined in the MQSINI configuration
file. The NSKSegidRange item in the AllQueueManagers stanza of the MQSINI
configuration file defines the range of Segment IDs used by MQSeries when
allocating memory segments.
When MQSeries is installed, this range is set, by default, to NSKSegidRange=10-20.
Since these memory segments are also allocated in the application program’s
process, you should ensure that applications do not allocate segments in this range.
You can force MQSeries to allocate its segments in a different range by editing the
NSKSegidRange item in your MQSINI file. After editing this item, you must stop
and restart all queue managers before the new value will take effect.
The range defined by the NSKSegidRange parameter must be wholly contained
within the Guardian Segment ID limits of 0-1023. The NSKSegidRange parameter
must define a range containing at least 10 Segment IDs for use by MQSeries.
Regards |
|
| Back to top |
|
|
| AndyMQ |
| Posted: Tue Nov 15, 2005 9:57 am Post subject: |
|
|
Apprentice
Joined: 22 Apr 2004
Posts: 33
Location: Scotland
|
Looks remarkably similar. I suspect the forty-eight might be just a coincidence and not a Guardian forty-eight
We have an identical queue manager and an identical application on another machine, and we don't have any problem with the segid range.
Found a possible small clue in the ZMQSSYS.README file which suggests using PARAM MQNEWSEGALLOC ON at the TACL level, although it won't do us any good as it's a pathway server, but we are gonna try adding this param to the server config and see if this works (that's once we get the appropriate password from our security folks).
Thanks. |
|
| Back to top |
|
|
| AndyMQ |
| Posted: Wed Nov 16, 2005 5:17 am Post subject: |
|
|
Apprentice
Joined: 22 Apr 2004
Posts: 33
Location: Scotland
|
Problem solved. The forty-eight was indeed a Guardian error. The CCSIDMEM file had security vector NUNU, but there was a subvol acl on at safeguard level which overrides the Guardian security. Unfortunately, when the queue manager was ported it was placed on a different disk, so the Safeguard settings applied were incorrect (d'oh!).
Just ran safecom, altered the subvol setting for the "D" subvol, and Bob's yer uncle!
Now the tricky bit, explaining it all to our security people.... |
|
| 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
|
|
|
|
