DPA and MWI to 2nd Call Manager cluster

clotheyes · ‎06-19-2008

Help!

I have an Octel voicemail system integrated with our call manager cluster (which works fine) which is to be connected to a 2nd Call Manager cluster which will continue to use the Octel for voicemail. I have the call routing correct such that the 2nd cluster can deliver calls to voicemail (phone on cluster 2 -> ICT to cluster 1 -> QSIG to Nortel -> Octel system). I also have the MWI signalling coming back from the Octel via the DPA. However, the MWI does not cross the ICT (as per the documentation which states that it shouldn't). I have tried linking the 2 clusters with a QSIG trunk (on a 6608 with a crossover cable between two E1 ports) to no avail. I'm not completely surprised by this, however we have another system (BT ITS) connected via a QSIG trunk (off an MGCP controlled IOS gateway) which tandems it's voicemail and MWI through the first ccm cluster and this works.

I can see in the CCM logs on both clusters that the MWI call is coming into the first cluster and being routed to the second cluster. The second cluster accepts the call but doesnt light the lamp (although I can see the call coming in to this cluster).

Any ideas on howto make this work (without using a second DPA as I have a third cluster to integrate soon too!)

Attached is a diagram of the system with the MWI messaging flow.

Thanks in advance

Mark

gogasca · ‎06-19-2008

Hi Mark,

Thanks for the diagram and the detailed explanation, could you please attach detailed CCM traces for both clusters when MWI is being sent

When u see this second call coming using either ICT or Q.SIG E1...into the 2nd cluster, what are the calling party and called party info?

cn field should be extension in cluster 2 and called party MWI ON number.

Please let me know the scenario for the traces, callinfo, timestamps and ccm version

Thanks again

clotheyes · ‎06-20-2008

Smart person here in the office worked it out (not me!)

MWI uses the QualifiedCalledPartyNumber whereas the second cluster wasn't translating this number for MWI. This now appears to be working over an ICT (with QSIG tunnelling).

Thanks,

Mark.