H323 use of Call Preserve and MTP

[Mike Lydick]

I have a gateway running 15.1.3T2. We require SRST and have enabled 'call preserve' under 'voice service voip'. We have also set the 'Allow Peer to Preserve call' in the CM Cluster Service Parameter. Call that are put on hold for longer that 30 seconds get dropped. TAC has recommending setting the Use MTP option, this resolves the issue but is not a good solution because it forces all traffic back to the MTP source which gets lost in failover to SRST.

I have noticed I do not have 'no h225 timeout keepalive', could this cause this behavoir with calls that get put on Hold.

The MOH and the MTP are software resources from the Callmanager.

UCM 8.5.1

If we do not use the Call preserve, Call hold works as expected so the behavoir definately is related to the use of call preserve.

[kusatija]

Hi Mike!

I do not think that an MTP is a must in the call flow.

Can you please elaborate the call flow, and provide the following.

Detailed cucm trace from all nodes for a call that fails after 30 seconds of being put on hold(plz uncheck the MTP box fpr the call), we need to check what message we get after 30 seconds that kills the call.

Running config from h323 gateway.

Following debugs from the router for the same call.

deb voice ccapi inout

deb h225 asn1

deb h245 asn1

Calling/called party numbers, codec settings bw MOH server, phone and the gateway.

please run these debugs off production, as i am not sure about your call flow.

Also can share how TAC reached to the conclusion that MTP is needed for sure, that will save a lot of time.

HTH

Kunal