CallManager - Nortel Option 61c Connectivity

[MPB123]

Overview:

PSTN ---> Nortel Option 61c ----> T1 MGCP C2821 -----> CallManager

Issue:

1. At a random time interval (approximately once per 24 hours), CallManager subscribers are unavailable to place calls through the 2821 to the PSTN/Nortel devices. At the same time, Incoming calls are routed properly from the PSTN/Nortel to CallManager subscribers.

Troubleshooting:

1. Errors in the CallManager trace files indicate that CallManager is receiving the following error from the gateway:

"47 2F Resources unavailable, unspecified The requested channel or service is unavailable for an unknown reason. This might be a temporary problem."

2. Utilization of the PRI at the time is at a maximum of 12 calls. Typically, it is around 4-6. We are quite confident that there is not 23 calls in place at the time.

3. It doesn't appear to be a route loop. CDR records show no loop at the time. If there was a loop, we would expect that inbound calls would not be processed. There also does not seem to be any pattern between various times (i.e. the same user/phone/RP/etc).

4. We are seeing "Cause i = 0x81FF - Interworking error; unspecified" errors in the router debugs (approximately 1 every 30-60 minutes)


Thoughts:

1. CallManager is set to use Top Down for channel selection, while the Nortel is configured to use RRB. Could this be causing a collision? If so, we have not been able to replicate it by generating call volume, as both sides seem to intelligently select an available channel.

2. The Nortel box is set to use SL1 as the IFC. Thought is to change it to S100.

3. Since the last outage, We have upgraded to the latest 12.4 release. (It was previously 12.4.4, it is now 12.4.8)


Resolution:

1. The issue is not resolved until the CallManager service on the publisher has restarted. Restarting the CallManager service on the subscriber (where all phone devices are registered) does not seem to resolve the problem.


Note: We now have debug isdn q931 set to a syslog server. We hope to get more complete debug details on the gateway next occurence.

Any thoughts?

Thanks!!!

 

[jneiberger]

Is this a QSIG trunk or just a plain PRI? What level of code are you running on the 61C? What version of CCM?

 

[MPB123]

Thanks for the reply.

It is not a QSIG Trunk - plain PRI. It is Succession 3.0, and CallManager 4.2.

 

[jneiberger]

Have you thought about making it a QSIG trunk just to see if it decides to behave? And why are you running such new CCM code?? ;-)

 

[MPB123]

Unfortunately we do not have a license for QSIG on the Nortel System.

 

[jneiberger]

Bummer. That would be a better option, but it might not make a difference in this case.

It really is too bad you're running 4.2. I think 4.1(3)SR3a is a bit more stable. Do you have SR1 installed?

 

[MPB123]

Yes - SR1 is installed.

Debug on router shows:

Cause i = 0x81AF1A01 - Resource Unavailable, unspecified

This occurs on all outbound calls until CallManager service is restarted.

Nortel is set to RRB, where Cisco is set to top down. Is it possible that this is a channel conflict?

 

[MPB123]

UPDATE -

We have determined that the system fails when the call reference number (from the ISDN debug traces) reaches 0x8300 (768 calls). At this point, outbound calls to the Nortel 61c fail. When CallManager service is restarted, the reference number resets. This allows outbound calls to continue until it reaches 0x8300 again....

Has anyone seen this issue before?

Thanks!

 

[jneiberger]

I wonder if this is related to Bug ID CSCsd13088? The symptoms aren't quite the same but they sound pretty close. Try disabling inbound faststart on CCM and see if that resolves the problem.

John