| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Execution server instances restarts too often |
« View previous topic :: View next topic » |
| Author |
Message
|
| Dio |
 Posted: Fri May 12, 2006 12:32 pm Post subject: Execution server instances restarts too often |
 |
|
Novice
Joined: 18 Jan 2006
Posts: 11
Location: Moscow, Russia
|
In time of high workload execution server instances restarts every 1-3 minutes. This slows down the system. No errors in logs, only instance stop/start.
This happens when RTRecycleThreshhold is not set or is set to 5000 - 50000. And there is no need to care of memory leaks cause memory usage is low (~30Mb per instance)
There was no changes when setting number of exec. server instances from 2 to 4.
Configuration is single box 1 processor Win2k server.
Why it happens???
more info:
Workload in 3 heaviest hours is starting of about 7000 process instances in hour. Process are of 2 types:
1. Straighforward process of 3 automatic activities
2. More complex process 6-7 automatic activities, 3-5 noops as selectors and 1 manual activity in ~1% cases.
There is 2 instances of 1st type per each instance of 2nd type.
Data structure is complex (about 50-60 fields) but goes the same throuth whole process (mean container reusage). |
|
| Back to top |
|
 |
| hos |
| Posted: Sun May 14, 2006 10:49 pm Post subject: |
|
|
Chevalier
Joined: 03 Feb 2002
Posts: 470
|
Hi,
obviously you have played around with the RTRecycleThreshhold variable (why would you do this if you do not experience memory leaks?) My best guess is that this variable is still taking effect. Have you restarted all servers after changing this value? Have you completely removed this entry in your configuration ( see the Installation Guide how to completely remove profile variables via fmczchk )?
In other words: execution servers never stop without a reason. Either this variable is still in effect or your system has a severe error situation. |
|
| Back to top |
|
|
| Dio |
| Posted: Mon May 15, 2006 12:55 am Post subject: |
|
|
Novice
Joined: 18 Jan 2006
Posts: 11
Location: Moscow, Russia
|
When I saw this situation first RTRecycleThreshhold was not set (it is by default). When we asked IBM support they told as about this variable, so I decided to try to play around with it. I restarted Workflow service every time I changed this variable.
And now I sure there is another cause of such behaviour. What else can I check to find it? |
|
| Back to top |
|
|
| hos |
| Posted: Mon May 15, 2006 4:02 am Post subject: |
|
|
Chevalier
Joined: 03 Feb 2002
Posts: 470
|
OK, so your execution servers are facing an error situation.
Have you checked the logs in the runtime database with fmcautil?
Have you checked errors file(s) of MQ and db2diag.log?
Have you run fmczchk to ensure proper MQWF setup?
I guess that you are facing some kind of resources problem as you say that this situation happens under heavy system load only. You already seem to be in contact with the support team so they'll probably ask for more details to analyze this problem. |
|
| 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
|
|
|
|
