asked on

Probable memory leak.. logs hurting my brain.

About once a week our server decides to fall in a heap.

Event log has got

Event Type:      Error
Event Source:      Srv
Event Category:      None
Event ID:      2020
Date:            27/04/2009
Time:            10:56:14
User:            N/A
Computer:      BIGONE
Description:
The server was unable to allocate from the system paged pool because the pool was empty.

and I'm fairly sure there is some memory leak.

I have been following the advice given here

https://www.experts-exchange.com/questions/24237916/Memory-Leak.html

and have got my poolmon output (attached zip file).

Looking at the logs, It looks like the MntA is causing the problem (I may be wrong), the logs are hurting my brain and I need help.

ASKER CERTIFIED SOLUTION

Member_2_3684445

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

pippipmilk

ASKER

I have requested the hotfix, MntA is the mount manager I believe.

Member_2_3684445

I found some other resources that all point to the mount manager.

Also...

MORE INFORMATION Currently, the mount manager performs all pool allocations by using the MntA tag. Therefore, all leaks that are related to the mount manager may appear to be similar.

pippipmilk

ASKER

I have put the hotfix on and need to arrange a reboot.. will keep you posted, but it's about a week between fails.

Member_2_3684445

Hehe, this post isnt going anywhere. And helping others is still more important than a reference in my case ;-)

I hope it helps..

Rgrds,

bdesmond

I don't see the poolmon output?

MntA is a pooltag it sounds like.

Thanks,
Brian Desmond
Active Directory MVP

pippipmilk

ASKER

I have attched the poolmon ouput file again
Poolmon-OUTPUT.zip

bdesmond

OK so you put the QFE in, is there poolmon data from after that?

Thanks,
Brian Desmond
Active Directory MVP

pippipmilk

ASKER

I'm getting some..

pippipmilk

ASKER

Well it has been stable for two weeks now so I'll take the hotfix as having solved the problem. Cheers