Showing results for 
Search instead for 
Did you mean: 

Ingress calls to CVP arrving over intercluster link?

This document was generated from CDN thread

Created by: Andrew Fraser on 22-08-2008 03:18:20 PM
Hello all,

My customer has a CVP solution using SIP with CUP and Unified CM 6. They also have a separate Unified CM cluster for IPT users. So far so good.

They now have a requirement to have incoming contact centre calls routed via the IPT cluster. I.e.

TDM PBX- pri – mgcp (QSig) – IPT CUCM Cluster
UCCE CUCM Cluster ----CUP --- ICM / CVP

Questions are:

Q1) Is this a supported protocol flow for incoming calls to CVP>UCCE? I understand that it is supported for calls to arrive at CVP from Unified CM via an MGCP controlled GW (with caveats) but what if these calls come via a different cluster with a trunk between them.

Q2) If it is supported, are the recommendations for the trunk configuration. Direct H323, direct SIP or GK controlled H323?

Q3) Are there any specific caveats of this approach? e.g.
  • Would we need MTP if SIP trunk?
  • What about DTMF?

It is LAN all the way between the ingress GW, 2 clusters, CUP servers, CVP servers and VXML GWs...

Subject: Re: Ingress calls to CVP arrving over intercluster link?
Replied by: Paul Antinori on 25-08-2008 08:18:02 PM
Hi Andrew,

With this call flow, all the calls will look like IP originated callers to CVP. It will work, but you will be lacking the survivability application that using an ingress gateway provides. TDM external type transfers will not work either (TBCT, *8) if you want them.

You only need the MTP on the SIP trunk if you are doing a warm transfer with queing, and the agent completes the transfer while still in queue. Practically, we document using a dual SIP Trunk configuration for this if you need to use it.

About DTMF, rfc2833 should be used consistently on all the endpoints to avoid DTMF capability mismatches and unexpected MTP allocation.

Subject: Re: Ingress calls to CVP arrving over intercluster link?
Replied by: Andrew Fraser on 27-08-2008 04:17:23 PM
Thanks Paul - much appreciated. I have a couple of follow on questions here.

1) Is there any recommendation from a CCBU perspective on what protocol the trunk should be between the two Unified CM clusters? I.e. SIP/H323/GK controlled. Does it matter, even taking into account DTMF? I'd imagine a SIP trunk would be sensible as we have a SIP trunk to the CUP server being used directly afterwards by the contact centre cluster...?

Note that it is possible that SCCP IP phones in the IPT cluster will want to call into the contact centre too over the trunk link between the two clusters.

2) You mention that endpoints should be SIP. However in this case the ingress GW will be MGCP controlled and there is a possibility (as mentioned above) that callers could be SCCP phones on the other side of an intercluster trunk link...will this cause any issues for DTMF?

Thanks again,
CreatePlease to create content
Content for Community-Ad
August's Community Spotlight Awards
This widget could not be displayed.