Low Virtual Memory Alert starts every 60-70 days

[Kumar Vishal]

We have a Cluster with a Pub and 4 Subs. Every 60-70 days, we start getting a low virtual memeory alert on the Publisher.

The threshold is set to 30% on RTMT (which means that when Virtual memory becomes less than 30%, RTMT will start alerting). We have to reboot the Publisher to get rid of this alert.

CUCM version is 6.1.3.3190-1 .

The Publisher server stats are -

HW Platform       : 7835H
Processors        : 1
Type              : Intel(R) Xeon(TM) CPU 3.40GHz
CPU Speed         : 3400
Memory            : 2048 MBytes

Is there any Bug associated with this CUCM version that could be causing some kind of a memory leak?

There are 2300 phones in the cluster and none is registered to the Publisher.

[Kumar Vishal]

The alert looks like -

Machine Name: CCM-PUB
Severity: Critical
App ID: Cisco AMC Service
Message: : 1310: Aug 05 22:11:31.261 UTC : %CCM_RTMT-RTMT-2-RTMT-ERROR-ALERT: RTMT Alert
Name:LowAvailableVirtualMemory Detail: Available virtual memory below 30 Percent. tomcat (975 MB) uses most of the memory.
The alert is generated on Thu Aug 05 18:11:31 EDT 2010 on node 172.29.0.180. Cluster ID: Node ID:CCM-PUB

[Aaron Harrison]

Hi

Not sure of the specific bug, but I had one customer on 6.1(3) with very similar issues and upgraded them to 6.1(4) SU2 a few months back and all has been well for them since...

Regards

Aaron

[claudiuprisecaru]

Hi,

I have the same problem, a problem that appeared after the upgrade CUCM 7.1 to CUCM 8.0 (3). I attached the error that appears in RTM.

Thanks,

[claudiuprisecaru]

I tried the solution described in the following document but did not work. After restart the  CiscoTomcate Service alert remains.

http://www.cisco.com/en/US/products/sw/voicesw/ps556/products_tech_note09186a0080a9ede5.shtml#low

Thanks,