Author |
Message
|
skoesters |
Posted: Mon Jan 26, 2009 12:30 am Post subject: AMQ6037: not enough memory for Websphere MQ |
|
|
Acolyte
Joined: 08 Jun 2008 Posts: 73
|
Hi,
i tried to start about 10 QMGRs on my Server (MQ 7 on Red Hat 5).
starting 1 - 8 works fine. 9 and 10 throw an error which says
AMQ6037: not enough memory for Websphere MQ (this is a free translation from german to englisch).
user root and mqm have ulimit = unlimited. The Server has 8GB Memory and about 1GB in use.
i do not know how to fix this.
Hope you guys can help me.
regards
Sebastian |
|
Back to top |
|
 |
Vitor |
Posted: Mon Jan 26, 2009 1:00 am Post subject: Re: AMQ6037: not enough memory for Websphere MQ |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
skoesters wrote: |
i do not know how to fix this.
|
Run fewer queue managers?
Linux is not a platform I know a lot about, but I suspect you're running out of shared memory / semaphores / whatever the Linux equivalent is rather than actual memory. Double check the Quick Beginnings for your platform and make sure there are no other settings you've omitted as part of the install process.
I know that on Windows there's this kind of bizarre system limit, where you run out of memory if you spin up more than about 7 queue managers, no matter how much actual RAM you have on board. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
skoesters |
Posted: Mon Jan 26, 2009 1:07 am Post subject: |
|
|
Acolyte
Joined: 08 Jun 2008 Posts: 73
|
i need to run 10
the problem with "running out of shared memory" should be solved with setting "ulimit -c unlimited" but this does not help
It seems that there is another strange limit in the system |
|
Back to top |
|
 |
Vitor |
Posted: Mon Jan 26, 2009 1:15 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
skoesters wrote: |
i need to run 10  |
Then you've got a problem!
skoesters wrote: |
the problem with "running out of shared memory" should be solved with setting "ulimit -c unlimited" but this does not help |
I did think that, but thought it was worth a spin.
skoesters wrote: |
It seems that there is another strange limit in the system |
It also seems you'll need someone with more Linux knowledge than me to help you find it!
As a final role of my personal dice, see if the queue manager is throwing an FDC file to match the message. That might contain a clue to what the queue manager is unable to get enough of. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
skoesters |
Posted: Mon Jan 26, 2009 1:36 am Post subject: |
|
|
Acolyte
Joined: 08 Jun 2008 Posts: 73
|
the thing with the quick beginnings was a good hint.
this seems to fix the problem:
kernel.msgmni = 1024
kernel.shmmni = 4096
kernel.shmall = 2097152
kernel.sem = 500 256000 250 1024
fs.file-max = 32768
net.ipv4.tcp_keepalive_time = 300
thanks! |
|
Back to top |
|
 |
Vitor |
Posted: Mon Jan 26, 2009 1:38 am Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
skoesters wrote: |
the thing with the quick beginnings was a good hint. |
Well yay me!
And yay you for posting the solution for the benefit of the wider community.  _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
|