Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 
New Member

Voice and Frame configuration

We have the following scenario:

Frame-relay hub and spoke design: head office - Newtown - our router (3640) has ethernet connection to VoIP router on customer LAN, and has 13 PVCs configured to remote-sites (all on the same physical serial interface).

256K PVC from Newtown to Leicester: at Leicester, our router has again an ethernet connection to a VoIP router on the customer LAN.

I have been looking at the configuration examples given on CCO, and have a few questions.

We want to ensure that the Voice traffic recieves priority over the data traffic on the 256K PVC between Leicester and Newtown.

How are we best doing this prioritisation?

As our routers have no Voice configuration, they are receiving the voice traffic from the VoIP routers simply as UDP packets. If so - do we need to do RTP/LLQ prioritisation, or can we use straightforward priority-list specifying the udp port for the voice traffic, and giving it a high priority?

If we create a map-class for the voice PVC, and shape accordingly, do we also need to create map-classes for each of the other 12 PVC's?

From reading CCO it looks as though if we don't, then the other sites will be shaped according to the default CIR of 56Kbps.

I am also confused as to how to classify the voice traffic - is it 'VoIP', is this 'VoIP over FR' or is it 'VoFR'?

any help much appreciated!

Cisco Employee

Re: Voice and Frame configuration

This would be VoIP over FR. It is best to use LLQ for the voice PVC, but there is another issue you will find. With all the PVCs on one physical interface, you will see jitter created by large data packets on the non-voice PVCs. You may need to fragment all the PVCs on this physical interface to avoid this. T1 or greater speeds, you may be OK. If this is needed, you will need to create a voice class data for the data PVCs that jsut fragments nad a map class voice for the voice PVC which fragments and does LLQ.

Hope this helps

New Member

Re: Voice and Frame configuration

thanks for your reply.

A further question - the access link into Newtown is 2MB total, but most of the other 12 data PVCs to remote-locations have CIR's of 256Kbps and under. How do I decide the fragment-size for each of the data PVC's? Is it related to the CIR-size?

CreatePlease to create content