Author |
Message
|
MQ_Lover |
Posted: Tue Dec 17, 2013 4:56 am Post subject: Too many amzla00 process - Windows |
|
|
Acolyte
Joined: 15 Jul 2013 Posts: 67
|
Hi All,
I have recently build new servers with MQ 7.5.0.1 on Windows Server 2008 and see that there are too many amqzla00 process spawning and they reach a limit which makes MQ Queue Manager crashes, as I tried to google it found it's bad app leaving connection opened, but the strange bit is the same app on old servers with MQ 7.0.1.6 never reported this issue and it's surprising just beeter servers and MQ latest is the difference, I tend to not agree that app is same way configured as we do a complete fresh install of MQ same is for apps also could be they have missed something on new servers, but just in case is there something from MQ end I could do to keep a check on these processes, any suggestions please.
Thanks |
|
Back to top |
|
 |
bruce2359 |
Posted: Tue Dec 17, 2013 5:35 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
What error was logged? Copy the first 50 lines of the FDC, and post them here. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MQ_Lover |
Posted: Tue Dec 17, 2013 6:38 am Post subject: |
|
|
Acolyte
Joined: 15 Jul 2013 Posts: 67
|
Hi bruce2359,
Please see below
+-----------------------------------------------------------------------------+
| |
| WebSphere MQ First Failure Symptom Report |
| ========================================= |
| |
| Date/Time :- Tue December 17 2013 10:58:24 GMT Standard Time |
| UTC Time :- 1387277904.719000 |
| UTC Time Offset :- 0 (GMT Standard Time) |
| Host Name :- GBW07546 |
| Operating System :- Windows Server 2008 R2 Server Enterprise Edition, |
| Build 7601: SP1 |
| PIDS :- 5724H7220 |
| LVLS :- 7.5.0.1 |
| Product Long Name :- WebSphere MQ for Windows |
| Vendor :- IBM |
| Installation Path :- C:\Program Files (x86)\IBM\WebSphere MQ |
| Installation Name :- Installation1 (1) |
| Probe Id :- XY551007 |
| Application Name :- MQM |
| Component :- xlsSpinLockRequest |
| SCCS Info :- F:\build\slot1\p750_P\src\lib\cs\amqxlspa.c, |
| Line Number :- 248 |
| Build Date :- Mar 8 2013 |
| Build Level :- p750-001-130308 |
| Build Type :- IKAP - (Production) |
| UserID :- MQM |
| Process Name :- C:\Program Files (x86)\IBM\WebSphere MQ\bin\amqzla |
| a0.exe |
| Addressing mode :- 32-bit |
| Process :- 00007064 |
| Thread :- 00000005 |
| QueueManager :- DGBW0PPCAD |
| UserApp :- FALSE |
| ConnId(1) IPCC :- 33460 |
| ConnId(2) QM :- 550 |
| Last HQC :- 2.5.5-7240824 |
| Last HSHMEMB :- 1.1.1-7264076 |
| Major Errorcode :- xecL_W_LONG_LOCK_WAIT |
| Minor Errorcode :- OK |
| Probe Type :- MSGAMQ6150 |
| Probe Severity :- 3 |
| Probe Description :- AMQ6150: WebSphere MQ semaphore is busy. |
| FDCSequenceNumber :- 0 |
| |
+-----------------------------------------------------------------------------+ |
|
Back to top |
|
 |
MQ_Lover |
Posted: Tue Dec 17, 2013 6:43 am Post subject: |
|
|
Acolyte
Joined: 15 Jul 2013 Posts: 67
|
Also as I see in the Task Manager there a loads of process continouslly spawning as below with something called as a fip number
C:\Progrma Files\IBM\Websphere MQ\bin\amqzlaa0.exe -mQMGR -fip41
this fip number keeps growing and gets to 1000 also is the culprit basically |
|
Back to top |
|
 |
bruce2359 |
Posted: Tue Dec 17, 2013 7:04 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
Did you search for probe id XY551007 ? _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MQ_Lover |
Posted: Tue Dec 17, 2013 7:17 am Post subject: |
|
|
Acolyte
Joined: 15 Jul 2013 Posts: 67
|
Hi bruce2359 - yes I did, but there is hardly anything relevant out there with respect to 7.5 or moreover the issue I am seeing all I could find is below which is not relevant here as mine is MQ 7.5.0.1 and the FFST stack is different also
http://www-01.ibm.com/support/docview.wss?uid=swg1IC81798
Error description
FFSTs with probe id XY551007 for components xlsSpinLockRequest
and xcsFreeQuickCell are seen for the WebSphere MQ V7.1 channel
pooling process ( amqrmppa.exe ). The major errorcode reported
is xecL_W_LONG_LOCK_WAIT with probe description "AMQ6150:
WebSphere MQ semaphore is busy."
The FFST stack reports:
MQM Function Stack
rppDisconnectPool
xcsTerminate
xcsDisconnectSharedSubpool
xlsThreadTermination
xcsFreeQuickCell
xlsSpinLockRequest
xcsFFST
Local fix
Problem summary
PROBLEM SUMMARY:
****************************************************************
USERS AFFECTED:
WebSphere MQ v7.1 on Windows platform.
Platforms affected:
Windows
****************************************************************
PROBLEM SUMMARY:
The locking contention for the quickcell resources results in
this FDC.
Problem conclusion
The code is modified so that the quickcell resource usage is
optimized and hence the locking contention is reduced. This
problem is fixed through the internal defect 155891.1.
---------------------------------------------------------------
The fix is targeted for delivery in the following PTFs:
v7.1
Platform Fix Pack 7.1.0.2
-------- --------------------
Windows U200351 |
|
Back to top |
|
 |
bruce2359 |
Posted: Tue Dec 17, 2013 7:38 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
Time to open a PMR. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
MQ_Lover |
Posted: Tue Dec 17, 2013 8:03 am Post subject: |
|
|
Acolyte
Joined: 15 Jul 2013 Posts: 67
|
Yep, have done that now, lets see what they come back with. |
|
Back to top |
|
 |
|