Low Virtual Memroy and Swap Partition Alerts

keithknowles · ‎03-20-2008

Help Please.

I am getting the following alerts from RTMT.

On Thu Mar 20 08:48:50 MDT 2008 on node <publisher IP>

Available virtual memory below 30 Percent.

ipvmsd (839 MB) uses most of the memory.

On Thu Mar 20 09:08:21 MDT 2008 on node <publisher IP>

Swap Partition available disk space below 10 Percent.

LowSwapPartitionAvailableDiskSpace partition has 6 percent of available disk space.

This began shortly after restarting the RIS Data Collector Service on CallManager to fix a gateway registration problem. Thanks in advance for your help.

Keith

Tommer Catlin · ‎03-20-2008

Typical Windows 2000 memory leaks. You will need to schedule a reboot of the server to free up the virtual memory. There is no way to "reclaim" the virtual memory. You could start and stop certain services, but this may or may not free up the virtual memory.

Check your available disk space also. The paging file could be adjusted to run better.

Normally, I recommend customer schedule maint on their CUCM servers for this reason. Eventually CUCM, if running over 50 percent utilized will need to schedule reboots every couple weeks. This will free up the memory and make it run smoother.

IF you are running CDR reports alot, this could cause some memory use, etc.

hope this helps

keithknowles · ‎03-20-2008

I should have stated this was a CCM 5.1 system, no windows 2000 involved. But we have started collecting CDR from CCM recently and I imagine the person in charge of that would be running reports on it, perhaps that is causing the leaks?

These servers have not been rebooted since sometime in the beginning of January. I find it odd that a person would have to reboot their call processor every few weeks, but it wouldn't be the strangest things I've come across with this system.

Tommer Catlin · ‎03-20-2008

If you search the bug list for 5.1 and "memory" there a number of bug fixes for memory leaks and your version.

Here is one I found with CDR.

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCsh58483&from=summary

But search the Bug Toolkit and you will find many memory issues with 5.1. Hate to say it, but 5.x has seen better days. 6.x seems to much more stable.

Cheers!

DWAM_2 · ‎06-10-2008

Hello,

I'am interesting with this behaviour.

What did you do to resolve it ?

We are on CCM6.0 and have got follow messages :

On Thu Jun 05 03:07:27 CEST 2008 on node 192.168.100.1. Swap Partition available disk space below 10 Percent. LowSwapPartitionAvailableDiskSpace partition has 9 percent of available disk space.

and

Message Erreurs CCM6 From Tue May 27 16:42:10 CEST 2008 to Tue May 27 16:50:10 CEST 2008 on node 192.168.100.2, there are 1 ThreadCounterUpdateStopped alarm(s) and 0 ThreadCounterUpdateResumed alarm(s) received. On Tue May 27 16:49:34 CEST 2008, the last ThreadCounterUpdateStopped alarm generated: ThreadCounterUpdateStopped NumberOfProcesses : 122 NumberOfThreads : 5759 Reason : Total processes/threads have exceeded the maximum tasks [1200] ProcessWithMostInstances : Process [cmoninit] has instances [25] ProcessWithMostThreads : Process [acserver] has Thread Count [5081] AppID : Cisco RIS Data Collector ClusterID : StandAloneCluster NodeID : CCMS1

Best regards.

Tommer Catlin · ‎06-10-2008

Log into RTMT and check to see if your diskspace is low. if it is, you may some logs building up or CDR taking up all the disk space.

DWAM_2 · ‎06-16-2008

Thank you for your answer.

On publisher :

Swap used : 96%

Active partition : 67%

Boot : 2 %

Common : 32%

SharedMemory : 1%

Inactive : 1%

How to see where the swap is affected ?

Best regards.

keithknowles · ‎06-10-2008

We rebooted the affected server. Our issue was related to a memory leak in the CUCM software.

DWAM_2 · ‎06-16-2008

Ok, I think so that we will reboot the server if we can not resolve the behaviour.

Elapsed time : 224 Days ...

DWAM_2 · ‎07-01-2008

Hello,

I rebooted the server 9 days ago.

I have again the same message :

Message Erreurs CCM6 From Mon Jun 30 18:50:43 CEST 2008 to Tue Jul 01 10:54:44 CEST 2008 on node 192.168.1.2, there are 1 ThreadCounterUpdateStopped alarm(s) and 0 ThreadCounterUpdateResumed alarm(s) received. On Tue Jul 01 10:54:16 CEST 2008, the last ThreadCounterUpdateStopped alarm generated: ThreadCounterUpdateStopped NumberOfProcesses : 121 NumberOfThreads : 1712 Reason : Total processes/threads have exceeded the maximum tasks [1200] ProcessWithMostInstances : Process [cmoninit] has instances [25] ProcessWithMostThreads : Process [acserver] has Thread Count [942] AppID : Cisco RIS Data Collector ClusterID : StandAloneCluster NodeID : CCMS3

Is it critical ?

Best regards.