asterisk users - Feb 2012 - Problem while sending SIP NOTIFY via AMI in 1.8.10-rc2

Hi,

while testing asterisk 1.8.10-rc2 I stumbled across a weird behavior. I
want to notify a snom phone to reload its configuration. For this to
happen, I use the NOTIFY mechanism. I started the notify via AMI
command. Asterisk is bound to udp 25060, because all phones are
registered with a local opensips proxy which uses 5060. The expected
behavior would be:
asterisk send SIP NOTIFY to the proxy, the proxy sends it to the phone.

Actually asterisk sends the packet to the proxy, but the contact header
contains something invalid (IMHO):

On Manager Interface:
        T 127.0.0.1:57530 -> 127.0.0.1:5038 [AP]
        Action: SIPnotify.
        Channel: SIP/max.
        Variable: Event=check-sync\;reboot=false.

Leads to:
        U 192.168.10.72:25060 -> 192.168.10.72:5060
        NOTIFY sip:max at 192.168.10.72 SIP/2.0.
        Via: SIP/2.0/UDP 192.168.10.72:0;branch=z9hG4bK1dff6efe.
        Max-Forwards: 70.
        From: "asterisk" <sip:asterisk at
192.168.10.72>;tag=as66766c2a.
        To: <sip:max at 192.168.10.72>.
        Contact: <sip:asterisk at 192.168.10.72:0>.
        Call-ID: 412a8eff76bd7ac56ac06831256fd6aa at 192.168.10.72.
        CSeq: 102 NOTIFY.
        Subscription-State: terminated.
        Event: check-sync;reboot=false.
        Content-Length: 0.

The weird thing is the port number 0 in the contact header.

Is this a bug or do I something wrong?

Thanks,

Karsten

Hi,

a little extension to my previous post: The phone sends "200 OK" for
the
NOTIFY via proxy to asterisk, but asterisk seems to ignore this. About
500 ms later, the NOTIFY is repeated by asterisk. This continues up to
the final timeout (with the typical log message).

Karsten