CUCM 10.6, Expressways C/E 8.5/MRA/Jabber/SIP (no H323). Got the whole shootin match configured (so ask me any questions besides this one) and working EXCEPT inbound from internet URI calls into Expressway-E. I have a soft-phone app called "Blink" thats registered to "sip2sip" on my computer. It is for URI internet dialing only, so you just put in something like firstname.lastname@example.org and it will query my DNS records and wind up (SUPPOSED TO) ringing my phones in my CUCM through the Expressways. It works perfect...in the sense that I see the perfectly constructed SIP invite coming into my Palo Alto destined for my Expressway-E outside interface. Then I see my Expressway-E immediately responding to the SIP invites source IP with ICMP. Why? What is that all about? Of course it never makes it into any call logs in the Expressways. It does this 6 times, one each for every inbound invite retry from the soft phone (that is not part of the CUCM at all). The ICMP response is "Destination unreachable/port unreachable". Why would the Expressway be wanting to ping 5060 after every incoming invite? Is it trying to 'verify' this is a legitimate inbound invite? I dont have any experience with public internet URI dialing/CUCM/Expressway until now. When I make a URI call from one of my registered Jabbers outbound to this SIP soft-phone over the internet (using Expressway outbound DNS search zone), it works perfect. This whole project was a blast to configure and learn, by the way.
Your Expressway-E outbound B2B are working fine?
Also have you tested inbound calls from any other source?
Also you are using Expressway with single or dual nic with NAT?
Check Collaboration Solutions Analyzer for B2B call test for inbound calls and collect the diagnostics logs on Expressway-E And put those logs in the log analysis of CSA and see what is the error you are getting.
Please rate if it is helpful and accept as a solution if applicable...
Well, I fixed it. I never thought Id get stung by something so simple. I see other guys write in when they have figured it out by something so simple, so I am too. I had the "UDP" protocol turned off under the Expressways configurations, protocols, SIP. I did this 'by the book' according to the Cisco config guides, and they just kind of skip over this little config part (configurations, protocols,SIP)and so did I, so I only had TCP and TLS on. Everything else worked fine, my MRA DX650 and 7811 phone are able to register to the Exp-E over the internet and so are my Jabbers. They all can call each other. I could successfully do an outbound URI dial to my little URI "Blink" soft-phone on a PC that is not part of the CUCM but I couldn't do a URI dial from that same "Blink" softphone into the Exp-E to one of my phones because it wants to use SIP 5060 UDP. What really threw me was the response from the Exp-E to the SIP 5060 UDP invite from the Blink soft-phone. Exp-E responds in Wire shark with ICMP/destination port not reachable to the source IP of the internet SIP Blink phone. Which in itself is true, the destination UDP port of 5060 was not reachable on the Exp-E but I was thinking it was telling me that UDP 5060 was not reachable on the SIP soft-phone that was trying to URI dial INTO my Exp-E. I am using the Exp-E with 2 separate NIC's and I do have a real static IP from Century Link the the EXP-E static NAT's to, just like in the Cisco config guides. Everything else worked perfect on this setup, so I thought, heck, its nothing I did, gotta be something somewhere else. No...
Thanks Anthony, that means a lot coming from you. Now I really, really know why Cisco skipped over the thought of even enabling SIP UDP 5060 in that one section in the Expressways, they did it to us for a reason, to stop us from enabling it and watching all the fraudulent call attempts come into the Expressway lighting it up like a switchboard at Christmas. Geeze.. Its never over.. onto CPL lists now.