Dropped Calls When All DAHDI Lines In Use

Home » Asterisk Users » Dropped Calls When All DAHDI Lines In Use
Asterisk Users 4 Comments

Hello,

I am running Asterisk 11.17 with DAHDI 2.9.0 and an OpenVox A800P with 8x analog POTS lines coming into my Asterisk server from the phone company. Internally, I
have about 180 SIP clients defined in sip.conf. What appears to be happening is that if existing calls are consuming all 8 external lines and a new SIP client attempts to make a call, an existing call gets dropped. The asterisk log simply shows this as a normal hangup, so I am not able to easily distinguish between a normal hangup and this type of dropped call. In testing, I am able to get a new SIP client to report “service unavailable” when all 8 lines are consumed, yet still drops are reported.

I have been unable to find any configuration settings pertaining to prioritizing existing calls over new calls. What else can I look for to attempt to debug and fix this so that existing calls are not dropped?

Thanks,

Andrew

4 thoughts on - Dropped Calls When All DAHDI Lines In Use

  • Have you given any thought to moving to at least a current supported version 13?
    Asterisk 11 has been EOL for some time now I doubt you will get a resolution to a version no longer supported. Moving to the latest version 13 should be relatively quick and painless, and if the issue persists you might find more assistance.

    John Novack

    Andrew Martin wrote:

    Dog is my Co-Pilot

  • —– Original Message —–

    John,

    Thanks for the reply. Yes, I am planning on moving to version 13 but need to find a solution in the interim. If there are any configuration options that pertain to which actions to take with existing calls when new calls come in, I think it is likely that they would be shared between both versions (and I want to make sure I have the correct settings when I switch to version 13 too). Can you advise on any tunables related to handling existing vs new calls?

    Thanks,

    Andrew

  • Andrew Martin wrote:
    I suggest you make the switch to the latest version 13, which should go fairly smoothly, and you may find that you no longer have an issue.

    JN

    Dog is my Co-pilot

  • You could use GROUP & GROUP_COUNT to track how many channels you are using before you attempt to dial out and send back a Busy/Congestion/Whatever to your endpoint when you are at your limit.