Paging and Intercom Seems Broken FreePBX 15


(Bifur) #1

Having an issue with a page group that pages several extensions (~15 or so). It is the first time I have came across the issue, but I have it happening to two separate installations.

Both servers are on Freepbx 15 and up to date on the stable branch. All extensions are using PJSIP. Server is cloud hosted. No issues with calls, etc. Phones are Yealink T46S

Here is a forum from Sept 2020 explaining part of the issue:
https://community.freepbx.org/t/page-group-calling-the-extension-initiating-the-page/70273

The bigger issue I am having is not the intercom calling the extension that is making the page, but this:
Periodically when you press the page speed dial (299 in this case) the intercom connects to the phones , blf lights turn red, but no audio is passing. If the phone is hung up then the active calls goes to holding and all of the phones are still connected. Hitting resume on the phone connects the person back to the call and then they can speak and everyone can hear them.

Second issue as the other link says, is that if 201 makes a page to page group including 201, 202, 203 then the page actually calls 201 as they are making the page. In past versions this did not happen and the one making a page would be skipped. As I said earlier, I have 2 different installs with this issue when making pages to page groups.

Has anyone came across this?

EDIT: Here is a pastebin of an example. It was to long to post here directly.

Edit2: https://issues.freepbx.org/secure/attachment/50005/page_example_FPBX15.txt
Cleaner call logs demonstrating issue.


(Bifur) #2

@cyberchaplain - I have a similar if not the same issue you had back in 9/2020. Did you find out anything about it? Thanks for any insight!


(Bifur) #3

I created another system and recreated the issue. Surely someone has came across this? I have submitted a bug report.

https://issues.freepbx.org/browse/FREEPBX-22463


(Bifur) #4

Update for anyone with this issue:

Changing the extensions to chan_sip makes paging work correctly. It does not call the extension making the page (or leave a missed call since it was not called).


(Brian Ladd) #5

Reverting to chan_sip is not a real solution, since chan_pjsip is the future and chan_sip will eventually be deprecated.


(Bifur) #6

I agree completely, but it was my solution for 2 systems that page heavily. Hopefully a bug is found and corrected. It works fine on FreePBX 14 using PJSIP.


(Dave Burgess) #7

Logs and Pastebin of a working session and a non-working session tied to a ticket would probably go a LONG way to helping out.


(Bifur) #8

I can post a working PJSIP example from FreePBX 14 but I wasn’t sure if that would be beneficial since they are different versions. I do not have a working example to post for FreePBX 15 using PJSIP.


(Dave Burgess) #9

The log from /var/log/asterisk/full for the non-working example is part of what we’d be looking for. That should give us the process the system is following as well as the error that’s being generated to keep the system from working.


(Jared Busch) #10
  1. Do you even have your page group setup correctly?
    image

  2. 15 extensions is a lot. You should consider multicast paging direct from the devices and not through the PBX.


(Jared Busch) #11

Since I have two phones on my desk, and everything is PJSIP, I tested this.

It is 100% broke.

Here is a the call log https://pastebin.freepbx.org/view/bb6b56cf

Here is how I have my page group setup.
image

I did see the incoming call on the phone I dialed from
image

I have a missed call.
image

From myself.
image

Yes, I have multiple AOR in use for ext 103.
image

In fact SangomaConnect also “rang” for the inbound page. That is a call flow I bet no one thought about…
image

Calling from extension 119 which has only a single AOR still results in a call to itself when dialing 179 from that extension.

Worse it actually autoanswered itself (this phone is mostly a defualt config so that could likely be fixed in a setting).
image

Edit: ticket updated with comments pointing to this post.


(Jared Busch) #12

A fixed was pushed to edge last night.

I’ll be able to test later this morning.