Internal calls are not connecting randomly

[JuanAvaya]

Hi all,
Since we added around 80 new H.323 phones to the system, we are getting more complaints stating users dial an internal number and do not get ringback. However, it rings on the other side but when they pick up there is only silence. Redialing is the solution here.

We are migrating slowly from Nortel to Avaya and as far as I was told, all failing calls are from Avaya to Nortel. I do not see any problem on the ISDN trunk between both systems. I have put a sniffer trace to one of the reported phones but had no luck so far.

Has anybody faced an issue like this one?

Regards

 

[kyle555]

What subnet/network region are the new phones in? What region are the trunks that connect to them in?

You can configure a gateway in 192.168.1.0 and phones in 192.168.2.0 and even if no IP routing exists between those subnets, if you put them in the same network region in the PBX, it will tell phones and gateways in those subnets to send packets to one another even though the network can't route them. That's how I'd start to address the problem - check between all the subnets in the network regions concerned. Station traces would show ******* for pktloss every 10 seconds if the phone or gateway lost all the packets.

 

[dialtonebone]

Any alarms? DSP Resource counts are good?
Kyle mentioned checking subnet masks - that's always a doozy. Check your IP phones and media gateways and make sure it all looks good on the network side, then start exploring your DSPs and IP Network Region setups.


Check failed calls with a list trace station and compare them to good calls.

 

[JuanAvaya]

We have a single NR for the whole system (It's quite small with only 290 phones). Subnet-masks are fine as well as the network side.
List trace shows all the same between a good and a bad call except for the voice info, where it is absent in the bad one. DSP resources are always under 30% of usage and have no alarms nor errors.

Avaya says sniffer trace shows all "ffffffffffffff" at Payload coming from Nortel and that the issue is on their side.

We are making some tests routing the same calls through a different trunk to the Nortel and see what happens.