[cisco-voip] 737730 Alarms kCtiMYTCPSendError

[Wes Sisk]

this is coming from CTIManager, so it would be caused by one of your 
endpoints establishing a TCP session to CTI Manager.  The primary 
communication to CTIManager from endpoints occurs on 
endpoint:random<->CMServer:TCP:2748

There are a couple of other connections for change notification and SDL 
links, but those should be to either local loopback address or other 
CallManager Nodes.  TCPView from sysinternals will show you all the 
connections to/from ctimanager.

Looks like the template for that message is:

Error: kCtiMYTCPSendError - MYTCP_Send: send error.
  App ID: Cisco CTIManager
  Cluster ID: MYCM-Cluster
  Node ID: MYCM04
  CTI Application ID: JTAPI[192]@CM01
  Process ID: 1308
  Process Name: CtiHandler
Explanation: TCP socket connection has been broken.
Recommended Action: Network error may have occurred. If problem 
persists, contact TAC..

All hits I see says this usually coincides with high CPU utilization by 
CTIManager.  Also looks like most folks restart CTIManager to get around 
the issue.  However, I do not see a good clear root cause and permanent fix.

/Wes


Terry Glass wrote:
> Have received over 73,7730 of the following alarms on a Publisher 
> running CCM 3.3(3)sr4a
> EventID: 0xE0000003(3)
> Error: kCtiMYTCPSendError - MYTCP_Send: send error
> Explanation: TCP socket connection has been broken
> Recommended Action: Network error may have occurred. If problem 
> persists, contact TAC
>  
> Any ideas where to start looking in the network?
> Thanks
>  
>  
>  
>  
> ------------------------------------------------------------------------
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>