Hi, after a deep refresh in TCP connections I finally worked out the issue preventing Snom phone to work reliably when using tcp. I had a look in user list and realized heaps of messages are exactly suffering this issue. The facts are: - Snom phones are configurable in the parameter "network id port" that defaults to blank in the phone web interface and in sipxconfig too. Default to blank, in Snom speaking, mean the port will be random! - Snom phones when exchanging sip messages via TCP use a Via header like this: Via: SIP/2.0/TCP 192.168.0.107:5060;branch=z9hG4bK-n7vfke23xhv0;rport So all communications initiated by the phone work great, cause the proxy will see the rport and reply to the correct port all the times. - All SIP messages sent by the proxy using TCP are failing cause port 5060 is not of course the random port chosen by the phone.
Setting network id port = 5060 in the phone web interface or in sipxconfig will make the port static thus resolving the issue. I'm setting 5060 as the default value in the Snom plug-in. So I'm asking you, do you think this is a non RFC compliant behavior? Forgive me if there is something not correct in SIP speaking ... I'm a real ignorant. Thanks in advance and best regards Alberto _______________________________________________ sipx-dev mailing list [email protected] List Archive: http://list.sipfoundry.org/archive/sipx-dev Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
