| Author |
Message
|
| SOLOHERO |
 Posted: Thu Feb 04, 2010 4:13 pm Post subject: Multiple Execution Groups |
 |
|
Centurion
Joined: 01 Feb 2007
Posts: 107
|
Hi There,
I have a problem..
Message Broker is using STDOUT for the n number execution groups. This is filling /var/mqsi.
So i want to remove the STDOUT seetings for the n execution groups..could some one tell me how do i do this..
Thanks
_________________
Thanks |
|
| Back to top |
|
 |
| smdavies99 |
| Posted: Thu Feb 04, 2010 11:02 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
can you tell us what files are growing large? an
Of the directory where these large files are may help
Also, how big is your /var/mqsi?
the output from the folloewing would help
The above assumes that /var/mqsi is on the same filesytem as /var.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| SOLOHERO |
| Posted: Sun Feb 07, 2010 7:54 pm Post subject: |
|
|
Centurion
Joined: 01 Feb 2007
Posts: 107
|
HI ,
Thanks for the reply..this prod and i dont have sufficient access to do the checks , but it is getting filled by some log files i guess...
what is the best way..i never heard abt this stdout..stuff
Any help is appriciated
_________________
Thanks |
|
| Back to top |
|
|
| Gaya3 |
| Posted: Sun Feb 07, 2010 10:11 pm Post subject: |
|
|
Jedi
Joined: 12 Sep 2006
Posts: 2493
Location: Boston, US
|
| SOLOHERO wrote: |
HI ,
Thanks for the reply..this prod and i dont have sufficient access to do the checks , but it is getting filled by some log files i guess...
what is the best way..i never heard abt this stdout..stuff
Any help is appriciated |
if you dont have the access talk to the respective team [Prod team]
Unix Admins might be knowing the best way to resolve this issue.
_________________
Regards
Gayathri
-----------------------------------------------
Do Something Before you Die |
|
| Back to top |
|
|
| jugo |
| Posted: Mon Feb 08, 2010 2:58 am Post subject: |
|
|
Novice
Joined: 17 Nov 2009
Posts: 13
|
It is worth understanding which files are filling up /var/mqsi because:
a) It might not be the STDOUT and STDERR files
b) Even if it is the STDOUT and STDERR files, the fact that they are filling up might be a warning of another problem. It would be judicious to check what is actually ending up in these files. |
|
| Back to top |
|
|
| Vitor |
| Posted: Mon Feb 08, 2010 5:08 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| SOLOHERO wrote: |
i never heard abt this stdout..stuff
Any help is appriciated |
Speak to the Unix admins, or any qualified Unix person on your site. They'll give you all the information you need on this stdin, stdout & stderr stuff. 
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| Vitor |
| Posted: Mon Feb 08, 2010 5:11 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
Which of course leads to another interesting question:
If you're unfamiliar with this stdout stuff (and why should you be familiar with it?) then how did you reach the conclusion that's where these files are coming from? Especially with limited access to production?
Is this another case of the Unix sys admins diagnosing problems by guesswork and finger pointing? 
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| nathanw |
| Posted: Mon Feb 08, 2010 5:18 am Post subject: |
|
|
Knight
Joined: 14 Jul 2004
Posts: 550
|
| Vitor wrote: |
Is this another case of the Unix sys admins diagnosing problems by guesswork and finger pointing? |
I am tempted to agree with Vitor here.
Unfotunately the finger of blame tends to hit middleware first no matter where I go
Recently i have been asked to invetsigate a few problems and out of the 5 4 were attributable to other problems, but as most on here know
That's Life
----
If i were you and as ou do not have the access to see the problems I owuld sit on the admins shoulder as he drives the desk and shows you the problem or get him to send you the logs etc so that we can try and diagnose the issue here
_________________
Who is General Failure and why is he reading my hard drive?
Artificial Intelligence stands no chance against Natural Stupidity.
Only the User Trace Speaks The Truth  |
|
| Back to top |
|
|
| meaton78 |
| Posted: Tue Mar 02, 2010 11:55 am Post subject: |
|
|
Centurion
Joined: 16 Oct 2008
Posts: 100
|
| We just recently started having this same problem. Our /var/mqsi filesystem is only 64mb, so it is filling up fast. I have verified that stdout is the culprit in a directory under /var/mqsi/components/broker/randomDirectory. I've looked at the file and it appears to be logging xml requests. I'm not quite sure what changed to start causing this. Any clues would be appreciated. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Mar 02, 2010 12:38 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
Can you post some of the output here. This would enable others to see what you are referring to and give advice
64Mb for a Filesystem? Wow that is small. I can't remember when I created a filesystem that small
Oh yes. It was in the days of 2.1Gb SCSI drives on an Alphaserver 2100 system.
Is there any logical reason why it is that small?
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| meaton78 |
| Posted: Tue Mar 02, 2010 12:47 pm Post subject: |
|
|
Centurion
Joined: 16 Oct 2008
Posts: 100
|
There isn't really anything that I can paste into a public forum. It appears to be all debug info from our HTTP execution group. I'm just not sure how any of that is set, since we have our broker log set to a different location.
I knew I'd get blasted about the file system size, but there really wasn't a need to give it a crazy amount of space since we knew what the size would be. Until recently, it was only about 15mb in use. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Mar 02, 2010 10:55 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| meaton78 wrote: |
| There isn't really anything that I can paste into a public forum. |
Why not?
Surely you can redact all sensitive data (eg broker names, ip addresses, usernames etc)?
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| meaton78 |
| Posted: Wed Mar 03, 2010 3:06 am Post subject: |
|
|
Centurion
Joined: 16 Oct 2008
Posts: 100
|
It's really just filled with xml that contains customer data. The only thing I can paste is as follows, but the flow that this refers to hasn't been updated in months:
- BKR.HTTP.SOAPRsFlow.BrokerWebService
- Applicable path o* for SOAPRsFlow.BrokerWebService
- Logging SOAPRsFlow.BrokerWebService to LOG.DIAGNOSTIC.MSGS
- Expiry disabled
- Set Format to XML
- Format disabled
- Persistence disabled
There was a deployment of new code Sunday night to this EG, but I'm not sure if there is a way for one flow to enable logging in some way for an entire EG. |
|
| Back to top |
|
|
| AndreasMartens |
| Posted: Wed Mar 03, 2010 5:00 am Post subject: Raise a PMR |
|
|
Acolyte
Joined: 30 Jan 2006
Posts: 65
Location: Hursley, UK
|
| To ensure you get a decent resolution I'd suggest you raise a PMR for this. |
|
| Back to top |
|
|
| joebuckeye |
| Posted: Thu Mar 04, 2010 7:07 am Post subject: |
|
|
Partisan
Joined: 24 Aug 2007
Posts: 364
Location: Columbus, OH
|
If the recently deployed code has Java Compute nodes in it make sure that:
a) there are no System.out.print() calls in there - this will be put into the stdout file
b) that if you use log4j or some other logging API that you are not dumping messages to the console as the stdout file is where all this goes, you may need to update the log4j.properties file (or it's equivalent for another logging API) to not dump to the console
If the new code is not using Java Compute nodes then I might contact IBM about this. |
|
| Back to top |
|
|
|
|
