| Author |
Message
|
| jmonte |
 Posted: Thu Dec 08, 2005 2:10 pm Post subject: Websphere MQ 5.3 on z/OS process CHINIT high % CPU |
 |
|
Newbie
Joined: 08 Dec 2005
Posts: 1
|
hi,
on z/OS 1.4 I have Websphere MQ 5.3.1 with high consumption of CPU by means of process CHINIT, consumes approximately 20 to 50 % of CPU, have had a similar problem !!!???
thanks
JMONTE |
|
| Back to top |
|
 |
| kevinf2349 |
| Posted: Thu Dec 08, 2005 8:55 pm Post subject: |
|
|
Grand Master
Joined: 28 Feb 2003
Posts: 1311
Location: USA
|
This may be better in the mainframe & CICS forum but....
When you CHINIT is taking up the CPU do you mean the CHIN address space?
What messages are you getting in the SYSLOG? Any? Are the address spaces defined correctly to WLM? Are they swapable or non-swapable
We have z/OS 1.6 and MQ 5.3.1 without any issues.
How many concurrently active channels do you have ? Is it one specific channel? What is it defined as, what values do you have for the channel attributes? Any firewalls involved? How busy is the network? Could be lots of things....more information would be nice  |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Thu Dec 08, 2005 9:40 pm Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
IF your connections are client connections via the Client Attachment Facility talk to IBM. There are some measures you can take...However the best is to optimize the client code. Opening and closing the Channel is very cost extensive.
Enjoy  |
|
| Back to top |
|
|
| hopsala |
| Posted: Thu Dec 08, 2005 11:45 pm Post subject: |
|
|
Guardian
Joined: 24 Sep 2004
Posts: 960
|
Hm, some possibilities, off the top of my head:
1. You left some trace running (WMQ trace, I mean)
2. Using SMF record in too short intervals.
3. LOGLOAD value is too low, thus a checkpoint is issued every other second.
4. A bug - what is your CSD level?
5. Possibily a very short DISCINT on channels, causing them to rapidly go up and down all the time.
6. A one milisec HBINT
7. Some channel exit for encryption, compression, or any other CPU-oriented function.
Try, using Omegamon or some other system facility, delving deeper into the CHIN address space to see which task is the culprit. But as kevin said, more info would be beneficial.
keep us posted  |
|
| Back to top |
|
|
| javagate |
| Posted: Fri Dec 09, 2005 4:25 am Post subject: |
|
|
Disciple
Joined: 15 Nov 2004
Posts: 159
|
I had a simular problem where I needed maintenance. How current are you with maint (what csd level)? However, this was years ago.
_________________
WebSphere Application Server 7.0 z/OS &
MQ 6.0. I work with WebSphere in the real world not in some IBM lab. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Fri Dec 09, 2005 8:15 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
| hopsala wrote: |
6. A one milisec HBINT
|
HBINT is specified in seconds....
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| hopsala |
| Posted: Sat Dec 10, 2005 4:00 am Post subject: |
|
|
Guardian
Joined: 24 Sep 2004
Posts: 960
|
| PeterPotkay wrote: |
| hopsala wrote: |
6. A one milisec HBINT
|
HBINT is specified in seconds.... |
Oh well, I did say it was off the top of my head  |
|
| Back to top |
|
|
|
|
