NEC & Call Manager Qsig

[kartboy]

Has anyone ever come across stange Ip addresses on the LAN interfaces of an NEC PABX when it also has connectivity to a Call Manager.

A recent event has caused a PABX to reset and was diagnosed as an abnormal loop between the Call Manager and NEC box over the Qsig links (4).

Call records dont show any volume of calls between the same numbers yet this has now happened twice and the LAN interface (not connected at all to the Call Manager) shows odd Ip addresses which are not part of our network Ip Adressing scheme.

Anyone have any ideas or has seen this before?

 

[whykap]

I've seen this with a call manager and NEAX 2200? PBX.

It had nothing to do with the LAN interfaces as I don't see how that is possible but with the QSIG trunks as your error codes pointed.

The issue was specific Directory numbers where forwarded from the NEC to CISCO and the same DN's routed right back from CISCO to NEC causing routing loops.

In this case call manager can handle the load but obviously the NEC could not causing high CPU usage and the PBX to lock up and eventually reload. Needless to say that with 1300 extensions on the NEC and several PRI's it took close to 45 minutes to come up 100 percent.

So check your call routing on both sides as it is probably the cause and it will happen again.

 

[kartboy]

Thanks, have started to look in that direction but its like a needle in a haystack, could any other traffic case do that or would it only be DN to a DN causing the routing loop?

 

[whykap]

DN, DID etc. it is a very easy mistake to do and since the cisco server has more processing power than the NEC the NEC ends up taking a dump.

If you debug your gateways on the CISCO side you might be able to see what is happening.
I agree with you it is very hard to troubleshoot.