Alerting on Partial Service/Out of Service

[Paul McGurn]

Can anyone tell me if/how it's possible to alert on when a UCCX service goes from "In Service" to "Partial Service" or "Out of Service'?  I find, that random events in our environment can cause services on UCCX to go into Partial Service, but I find it nearly impossible to find out "when" so I can narrow down a timeframe in the logging.  I'm hoping this can be done via RTMT, but I wanted to ask the community in the off chance someone has already set it up.

UCCX version 8.5.1.11003-32

[jason.mcdonald]

Hi Paul,

Yes it's definitely possible and in fact I believe simple snmp monitoring will do it for you. I know you said you have 8.x but the document below should be a good start - chapter 3 page 3-107 and chapter 4

http://www.cisco.com/en/US/docs/voice_ip_comm/uc_system/UC_Managed_Service/UCCX_Managed_Services/UCCXManagedServices701.pdf

I'll see if I can get you the exact trace levels but I'm pretty sure snmp setup and a says log server should do it for you.

Sent from Cisco Technical Support iPad App

[Paul McGurn]

It'd be great if you could let us know which trace setting powers that.  We already have SNMP trap monitoring set up, but we're definitely not getting traps for partial/OOS events.

[jason.mcdonald · Accepted solution]

Paul,

If you go to UCCX Serviceability and under alarm change the syslog setting to Informational (or higher) youll get something like this when JTAPI or another service goes into partial or out of service:

03-31-2013    21:22:34    Local7.Critical    10.0.200.21    648: Apr 01 2013 01:22:35.271 UTC :  %UC_SS_TEL-2-SS_PARTIAL_SERVICE: %[Failure reason=A number of CTI ports are OOS - TPG[id=0,state=PARTIAL_SERVICE]:Ports[1001]][App ID=Cisco Unified CCX Engine][Cluster ID=][Node ID=UCCX85]: JTAPI subsystem in partial service

03-31-2013    21:22:34    Local7.Info    10.0.200.21    647: Apr 01 2013 01:22:35.263 UTC :  %UC_SS_TEL-6-JTAPI_PROVIDER_EVENT: %[JTAPI provider name=10.0.200.19,Addr=1001][Event=CiscoAddrRemovedEv received][App ID=Cisco Unified CCX Engine][Cluster ID=][Node ID=UCCX85]: JTAPI Provider state is changed

Here is a screenshot of what I have setup:

And here is the exact alarm I pasted above:

And here is what youll see when it recovers:

Hope this helps you.

Jason

[jason.mcdonald]

Hi Paul,

Just wondering if this was what you're looking for or if you need more info?

Jason

Sent from Cisco Technical Support iPad App

[Paul McGurn]

Yes, looks to be.  I've been swamped for the past week, so I haven't had a chance to dig in and tweak anything on our end.  Thanks much!

[Erick Bergquist]

Does anyone know if there is a SNMP OID that can be polled to check for partial service, or HTTP request (REST, etc) similar to the http://serverIP/uccx/IsDBMaster check? 

I've looked through the SNMP MIB the Serviceability PDF mentions but not finding anything. 

The downside to the syslog method is if the system is restarted and the engine never gets to IN SERVICE, it will be partial and you may not get alerted via syslog since it wasn't fully IN SERVICE and went out of service, etc. 

Thanks, Erick