Receive video call from Spark client to external company via cisco Jabber B2B

[dsobrinho]

Hi guys,

Does anybody knows something like that or have the same experience problem? The log the call flow answer with 503 but I have ringback.

2018-03-14T17:21:02.974-03:00 tvcs: UTCTime="2018-03-14 20:21:02,969" Module="network.sip" Level="DEBUG": Action="Received" Local-ip="10.5.196.20" Local-port="25237" Src-ip="10.5.196.15" Src-port="5560" Msg-Hash="10160664254878059712"
SIPMSG:
|SIP/2.0 503 Service Unavailable
Via: SIP/2.0/TCP 10.5.196.20:5060;egress-zone=CUCMNeighbor;branch=z9hG4bK3e5f1aed8076952a37b900a3e2f58aa3173193.f669023973fd707dd2883904c93196df;proxy-call-id=6d75db8b-3749-4580-b717-d1a32aaeee49;rport,SIP/2.0/TLS 10.5.202.130:5061;egress-zone=DNSZoneB2B;branch=z9hG4bK1e239b26b806905d55dfd48aed6834a953114.cb7bfcbaa0a55bfdf1b65fd99bb20b44;proxy-call-id=f1adefc7-c420-49a4-a033-f149af96e25d;received=10.5.202.130;rport=25162;ingress-zone=DefaultZone,SIP/2.0/TLS l2sip-cfa-01.wbx2.com:5061;branch=z9hG4bKdf76c43c31d97bd624e8726583823a7b;received=18.221.216.175;rport=49908;ingress-zone=DefaultZone,SIP/2.0/TLS 127.0.0.1:24565;branch=z9hG4bK-323435-94853063c114aaa30e3b5a7c5838e2ce;rport=44473
Call-ID: 0c8308f9b9d99921ba6f095c1474a261@127.0.0.1
CSeq: 1 INVITE
From: "Daniel Rodrigues Anunciado Sobrinho" <sip:Daniel.Sobrinho@added.call.ciscospark.com>;tag=1826021116
To: <sip:suporte.added@acme.com.br>;tag=133618438~80f843ae-9636-4957-95d8-0e16860abb0e-21678678
Server: Cisco-CUCM11.5
Date: Wed, 14 Mar 2018 20:20:59 GMT
Allow-Events: presence
Reason: Q.850 ;cause=47
Session-ID: 00000d6200105000a00000059a3c7a00;remote=231f537a356ca467722196ec532ca0c8
Content-Length: 0

[Adam Pawlowski]

This may not be too helpful but I'm curious since we're setting a lot of this up now. I don't follow your call flow here. Are you using hybrid call connector to call as yourself from Spark via on-premise routing, and you're trying to dial back out to the Expressway? That's what it looks like - that cause code looks like it wanted to pull in some sort of MTP or something for some reason which it couldn't do and failed the call out. IIRC that was in the notes for why they moved to the Spark-RD instead of the CTI RD since those always sucked in a MTP of some sort. I could be completely off.