[RTMT-ALERT] LowCallManagerHeartbeatRate

pmaglinger · ‎07-07-2005

I get [RTMT-ALERT] LowCallManagerHeartbeatRate alerts, usually in the early morning hours, from my CallManager. The body of the message is:

At 01:05:32 on 07/07/2005 on node XXX.XXX.XXX.XXX.

Call Manager heartbeat rate below 24 beats per minute.

Current heartbeat rate is 23 beats per minute.

I've seen the heartbeat as low as 19 beats per minute. I've found very little documentation on this. Can you tell me what this means and if there should be concern.

thisisshanky · ‎07-07-2005

The heartbeat is nothing but a keepalive that RTMT sends to Callmanager service and if it doenst respond back that heartbeat is considered to be lost. Heartbeats could be lost due to higher CPU utilisation on the box, or if the CM is actively processing calls. (Like a subscriber) Do you have any third party Apps running on this server resulting in higher CPU?

Sankar Nair
UC Solutions Architect
Pacific Northwest | CDW
CCIE Collaboration #17135 Emeritus

pmaglinger · ‎07-07-2005

I don't believe so. About the only other thing running on it is Cisco Security Agent, and that's Cisco approved. I would skin somebody alive if they put any other kind of app on there.

pmaglinger · ‎07-07-2005

Another thing is that this is happening around 1 am. Very little call volume going on at that time. In fact, very little of anything is going on at that time.

travis-dennis_2 · ‎07-07-2005

Just to chime in I have seen this at a lot of my customer sites. Since 4.0 I have been seeing it more during periods of very low (usually none) call activity. No other process running (backup, AV, etc). For lack of desire to debug it I have attributed it to "sleepy server syndrome". Seems like if the server is doing absolutely nothing it does not respond to keepalives in a timely manner all the time. There seems to be no adverse implications that I have found but it is interesting nonetheless.

Hope this helps.

Please remember to rate all replies

paul.harrison · ‎07-08-2005

1AM, sounds like it's probably around the time that the Call manager backup runs.

It's usualy 2AM but allowing for daylight saving and clock errors, 1AM could well fall into the backup time.

Paul

pmaglinger · ‎07-08-2005

Could be, but thinking back this does seem to occur during the off peak hours. I had a couple of messages around 4am this morning, well after the backup period. I voting for the "sleepy server syndrome".

stevecrumbaugh · ‎10-25-2005

I have seen similar issues when backups run - this could spike the CPU and freeze out the heartbeats. Are you using an anti-virus program that could be updating? You may want to check the App Log (in event viewer) which will show other events that could contribute to this?

pmaglinger · ‎10-25-2005

Acutally after a lengthy bout with Cisco TAC, reviewing log files time and time again, we finally determined that it was caused by having a large call record and the time of day that CCM ran maintenance on the call record database. We needed to have 3 months of call records, so I just adjusted the RTMT threshold down so it wouldn't spew out emails.