[SOLVED] Call aborts after 15 minutes

Asterisk Asterisk Support
1

Using asterisk 13.11.2/PJSIP my calls abort after 15 minutes, when the peer sends an UPDATE message like this:

09:49:16.905139 IP (tos 0x80, ttl 251, id 0, offset 0, flags [none], proto UDP (17), length 575)
    62.52.148.38.5060 > 77.180.62.22.5060: SIP, length: 547
        UPDATE sip:LOCAL-NUMBER@77.180.62.22:5060 SIP/2.0
        Via: SIP/2.0/UDP 62.52.148.38:5060;branch=z9hG4bKjqtq0g10dgig579c9370smp0lutp3.1
        To: <sip:LOCAL-NUMBER@sip.alice-voip.de;user=phone>;tag=62df3df7-8f8e-40b5-a4e2-347c3c56f3c8
        From: <sip:REMOTE-NUMBER@sip.alice-voip.de;user=phone>;tag=669730326-1476171253436
        Call-ID: c670df90-4fd7-4d4c-b3ba-37bd765e56bc
        CSeq: 424992755 UPDATE
        Max-Forwards: 68
        Content-Length: 0
        Contact: <sip:REMOTE-NUMBER@62.52.148.38:5060;transport=udp>                                                        Supported: timer
        Min-SE: 120
        Session-Expires: 1800;refresher=uac

As asterisk does not respond to the above UPDATE message, the peer keeps resending the UPDATE an additional 10 times, followed by a BYE message that is also resent a number of times as there is no response from asterisk either.

I understand this has something to do with session timers, but what exactly is the problem here, why does asterisk not respond to the UPDATE? Is there something I needed to configure?

2

Turns out that this problem seems to be caused by states timing out on the firewall and the firewall rule that is supposed to let inbound udp traffic to port 5060 pass from my ITSP’s SIP server address being ineffective because of the A record pointed to in the DNS changing in a way I did not expect it to.