After some recent updates I lost all audio to remote extensions

Free PBX 13.0.191.11, Asterisk Version: 13.15.0

Over the weekend I applied all module updates and installed firmware 10.13.66-20

All of my extensions are remote using pjsip on port 51601, and were working last week.

I get no audio at all, from the logs it looks like asterisk is trying to talk to the remotes actual ip address, not the natted public. 192.168.2.215 is the actually ip address of the phone.

Any ideas?

Got RTP packet from 208.93.42.132:62758 (type 00, seq 000924, ts 124800, len 000160)
Sent RTP packet to 192.168.2.215:2236 (type 00, seq 012585, ts 124800, len 000160)
Got RTP packet from 208.93.42.132:62758 (type 00, seq 000925, ts 124960, len 000160)
Sent RTP packet to 192.168.2.215:2236 (type 00, seq 012586, ts 124960, len 000160)
Got RTP packet from 208.93.42.132:62758 (type 00, seq 000926, ts 125120, len 000160)
Sent RTP packet to 192.168.2.215:2236 (type 00, seq 012587, ts 125120, len 000160)
Got RTP packet from 208.93.42.132:62758 (type 00, seq 000927, ts 125280, len 000160)
Sent RTP packet to 192.168.2.215:2236 (type 00, seq 012588, ts 125280, len 000160)

I added a chansip extension to the same remote phone and I got audio going both ways. So, it seems the problem is with pjsip.

Got RTP packet from 208.93.226.13:12488 (type 00, seq 049926, ts 1733228518, len 000160)
Sent RTP packet to 75.99.xxx.xxx:2254 (type 00, seq 037750, ts 1733228512, len 000160)
Got RTP packet from 75.99.xxx.xxx:2254 (type 00, seq 018040, ts 2089945161, len 000160)
Sent RTP packet to 208.93.226.13:12488 (type 00, seq 021136, ts 2089945160, len 000160)
Got RTP packet from 208.93.226.13:12488 (type 00, seq 049927, ts 1733228678, len 000160)
Sent RTP packet to 75.99.xxx.xxx:2254 (type 00, seq 037751, ts 1733228672, len 000160)

Added a fresh remote pjsip extension and it’s trying to talk to the phones private network address…

lh*CLI> rtp set debug on
RTP Debugging Enabled
[2017-05-15 22:30:15] DEBUG[1202][C-00000038]: res_rtp_asterisk.c:4706 ast_rtp_read: RTP NAT: Got audio from other end. Now sending to address 162.212.245.118:16528
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000309, ts 049440, len 000160)
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000310, ts 049600, len 000160)
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000311, ts 049760, len 000160)
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000312, ts 049920, len 000160)
Sent RTP packet to 192.168.201.89:2270 (type 00, seq 016652, ts 049920, len 000160)
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000313, ts 050080, len 000160)
Sent RTP packet to 192.168.201.89:2270 (type 00, seq 016653, ts 050080, len 000160)
Got RTP packet from 162.212.245.118:16528 (type 00, seq 000314, ts 050240, len 000160)
Sent RTP packet to 192.168.201.89:2270 (type 00, seq 016654, ts 050240, len 000160)

On another PBX, exact same version… All pjsip internal extensions… I added a remote pjsip extension and I get the same problem. The pbx is trying to talk to the remote phone using it’s private network address…

Seems nat is broken after these new updates…

Got RTP packet from 208.93.42.132:16516 (type 00, seq 000939, ts 150240, len 000160)
Sent RTP packet to 192.168.201.89:2224 (type 00, seq 026631, ts 150240, len 000160)
Got RTP packet from 208.93.42.132:16516 (type 00, seq 000940, ts 150400, len 000160)
Sent RTP packet to 192.168.201.89:2224 (type 00, seq 026632, ts 150400, len 000160)
Got RTP packet from 208.93.42.132:16516 (type 00, seq 000941, ts 150560, len 000160)
Sent RTP packet to 192.168.201.89:2224 (type 00, seq 026633, ts 150560, len 000160)
Got RTP packet from 208.93.42.132:16516 (type 00, seq 000942, ts 150720, len 000160)
Sent RTP packet to 192.168.201.89:2224 (type 00, seq 026634, ts 150720, len 000160)

We are seeing the same issue with PJSIP. Not just natted endpoints, local endpoints showing the same problem on internal calls.

Had to roll-back the Asterisk version to fix in a pinch.
yum downgrade asterisk*13.14.1
reboot

1 Like

Should this be placed in the Bugs area so the developers will be sure to see it? And…

If using PJSIP, should we hold off installing this upgrade or be prepared to downgrade Asterisk?

Thanks.

John

If you want to be sure the developers see it, you need to submit it as an “issues” ticket.

Having said that, have you tried setting up a Chan-SIP connection using the same credentials to see if Chan-SIP works? If the problem goes away on Chan-SIP and is reproducible on PJ-SIP, then you have a smoking gun that you include on the ticket.

It’s possible that it might be a problem at the remote end and is simply a coincidence. Obviously, since several people are seeing it, it could be something in the PJ-SIP stack, but if it was, we’d be hearing a LOT of noise about this and it doesn’t seem like a widespread problem.

I did get this working last night by switching all my remote endpoints to chan_sip. I also tried connecting remotely to another pbx with pjsip and it failed. Both of thse devices are FreePBX Phone System 500s and I had applied all module updates and 10.13.66-20. I can test this on some other pbx’s I have out there, though it doesn’t look like a coincidence to me.

Same here, Im going to downgrade too.

Anyone solve this yet or know the cause? It seems an issue only with some configs.

There was a flurry of announcements here about a week (10-ish days) ago about a couple of bugs that made this (or something very much like this) happen. The newest version of the system should solve the problem.

Just tried the latest Asterisk yesterday and still the same symptoms. Most of my systems do not have an issue. There is not much different with the failing system. Things I can can think of: internal net is a /16 we have multiple routed 192.168.x.x subnets.

Just wanted to say that I’m having the same issue as described. I didn’t see anyone else’s post about this until now so I created another thread. I’m using PJSIP on my extensions, applied updates awhile ago, and did not see the issue until the Asterisk service was restarted.

There was a problem with freepbx core a couple of weeks ago and pjsip not setting the bind address or something make sure you are running the latest version of freepbx core by doing a full module update or search on here

I think its a separate issue from the core update. I did have that issue with internal calls on core v13.0.120.0 and I have tried on the latest module update (13.0.120.3). This seems to be related to when I restart asterisk and it completes its upgrade from 13.14.0 to 13.15.0.

Here is my issue that I submitted - Asterisk Reboot Breaks External Audio

I don’t know if this is related, but we have two extensions down at a branch office. They connect to FreePBX over a nailed router to router IPsec tunnel. We can ping them from the FreePBX box, but they have orange lights and sip show peers shows their status as unknown. Some extensions at the branch office are working, though.

We updated modules today about the time the problem started, but I don’t know if that is related or coincidental.

This continues to be an issue with asterisk 13.15 and 13.16. I am having to manually hold my systems at 13.14.*

Also, I installed the latest Freepbx 14 stable with asterisk 14 and the same issue is present with PJSIP extensions.

This seems like a very severe issue to me to have to hold back upgrades and especially given that the supposedly stable Freepbx 14 (SNG7-FPBX-64bit-1706-1) has this problem with a fresh install.

Can someone please work on correcting this problem?

So you are clear, this is an Asterisk problem, not a FreePBX issue. Posting here with pleas for someone to fix are misplaced. However, the issue is supposed to be addressed in Asterisk 13.17 and 14.6, although I haven’t yet tested to confirm this. Neither version is available yet in FreePBX 13 as they are very newly published and we’ve not yet completed testing internally.

Thanks for your reply. I agree that this is an Asterisk issue.

I just updated the FPBX14 machine to Asterisk 14.6 (it seems it came out today), and I can confirm that PJSIP now works. I will watch for 13.17 to see if that corrects the issue for that branch.

Asterisk 13.17 is available on the asterisk website. However, it is not available on the shmz repository.

Can you please push to have this added to that repository asap?

Thanks for your help!

Asterisk 13.17 has been out for 2 weeks now, but it still does not appear in the SHMZ repository.

When should we expect it to be there so that we can stop worrying about doing a yum update on any of relevant installation?