Disconnecting call for lack of RTP activity in 31 seconds

[Stewart1 (Stewart)]

Regarding pjsip and FPL, see

Add Trunk SIP (chan_sip) Trunk

Back in 2018, pjsip was not compatible with trunks hosted by Fibernetics, including FPL. The problem was that pjsip sends the 200 OK response to an incoming INVITE with a Contact header that has no username. This is perfectly valid according to the RFCs, but it somehow caused the Fibernetics server to not respond with an ACK, so incoming calls failed. There was a long thread about this, in which I suggested various ways of getting it fixed, but there were no further updates. AFAIK, pjsip st…

and the thread linked from there.

You can confirm with pjsip logger if the bug is still there, or if your issue is different.

For the call drop issue, the usual problem is a NAT association timeout. If that’s the case, you can confirm that with BYE from the remote end: Make a test call to your mobile. A few seconds after answering, hang up the mobile and confirm that within a second or two, the calling phone shows the disconnect. If not, we will troubleshoot that first. If the test succeeded, make another test call to your mobile, answer it and keep the connection up for 13 to 14 minutes. Now hang up the mobile and see whether the calling phone shows the disconnect. If not, you have a timeout issue.

Confirm that your trunk configuration has
qualify=yes
any SIP ALG in your router/firewall is disabled, you’re not doing any funny stuff like forwarding a different external port to the port that chan_sip Bind Port is set.

If you still have trouble, post router/firewall make/model. If it doesn’t have a public IP address on its WAN interface, post modem make/model and ISP.