PRI via TE120P just stops working Outbound

[bradcook]

Hi all,

Hoping someone out there might be able to help me ...

I have a PIAF 1.3 (32-bit Asterisk 1.4) build on a Dell Dimension workstation and a TE120P for terminating a PRI/10

Up until 3 weeks ago, it was working fine.

However, 6 days ago I ran update-scripts/update-fixes to be sure I was fully up-to-date, and lo my outbound calls stopped routing (all circuits are busy).

Snip of logfile :

[Oct 31 16:01:34] WARNING[3295] app_dial.c: Unable to create channel of type 'ZAP' (cause 0 - Unknown)
[Oct 31 16:01:34] VERBOSE[3295] logger.c: == Everyone is busy/congested at this time (1:0/0/1)
[Oct 31 16:01:34] DEBUG[3295] app_macro.c: Executed application: Dial
[Oct 31 16:01:34] VERBOSE[3295] logger.c: -- Executing [s@macro-dialout-trunk:20] Goto("SIP/876-08553578", "s-CHANUNAVAIL|1") in new stack
[Oct 31 16:01:34] VERBOSE[3295] logger.c: -- Goto (macro-dialout-trunk,s-CHANUNAVAIL,1)

Inbound calls on the PRI are fine.

I've gone back to basics and can see that my configs have not changed from the working versions.

My group=0 is consistent, from-pstn context is consistent, g0 is the only trunk defined and is not disabled, and all standard codecs are allowed. I've moved the PCI card to the only other PCI slot, with same result (no success).

Has anyone seen this kinda behaviour ? Can anyone suggest anything outside the "usual" fixes I've tried for PRI trunks not being available outbound ?

My status screen output is below.

Any suggestions would be welcome - I think I've been looking at this for too long

Regards,
Brad

Status Version 1.2.6 released on Date 100708
pbx.local on 192.168.0.204
********************************************************************
* PBX in a Flash Version 1.3 Daemon Status *
* Running Asterisk 1.4 *
********************************************************************
* Asterisk * ONLINE * Zaptel * ONLINE * MySQL * ONLINE *
* SSH * ONLINE * Apache * ONLINE * Iptables * ONLINE *
* Fail2ban * ONLINE * Ethernet0 * ONLINE * IP Connect * ONLINE *
* BlueTooth * ONLINE * Hidd * ONLINE * NTPD * ONLINE *
* Sendmail * ONLINE * Samba * ONLINE * Webmin * ONLINE *
********************************************************************
* Running Asterisk Version : Asterisk 1.4.21.2
* Asterisk Source Version : 1.4.21.2
* Zaptel Source Version : 1.4.12.1
* Libpri Source Version : 1.4.7
* Addons Source Version : 1.4.7
********************************************************************
CentOS release 5.2 (Final) - 32 Bit ** Kernel: 2.6.18-92.1.6.el5
root@pbx:/var/log/asterisk $

 

[bradcook]

Oh, sorry. Forgot to mention debug

"pri debug span 1" shows NO entires when I dial out. (shows plenty on successful inbound calls so debug is definitely enabled when I test it)

Pls note the PRI is currently plugged into a Trixbox 2.3 system that is the "production" pbx, using same (as far as I can tell) configs, so we know the PRI is a working one.

I'd like to avoid a rebuild if possible, coz haven't got clean "whole system" restores working yet and not keen on retyping a 50-ext config, but will go there if needs be.

Also happy to consider the Asterisk 1.4 v 1.6 question, if it gets us a working pbx.

Very keen to get PIAF working, because Trixbox is just not for us anymore.

Cheers,
Brad

 

[bradcook]

Thanks all for the input

Fixed now after a paid consultancy.

Regards
Brad

 

[euser4life]

Do you mind sharing what the resolution is? It might help someone else down the road.

 

[bradcook]

Yes of course. Sorry for the lack in my earlier post. Just wanted to flag it as solved so no-one wasted time on it.

I had intended to do post details later today, after getting on with my cutover now, but I'm in a holding pattern on that so here goes ...

Okay - basic issue was what appears to have just been a bad PiaF install. Can't put a finger on the "bad" bit, sad to say.

It all "looked" okay in terms of configs & log output, but the ISDN behaviour as described was just weird.

Odd things happened, like how zap was dealing with leading 0s in long distance dial strings. These are required by my carrier, but on this build suddenly they were reported in PRI debug as not allowed.

Consultancy worked with me as a fresh set of eyes. Without any great revelations, and in confirming that what I'd done seemed okay & should have worked (phew for my sanity), it helped me to bite the bullet and take the backup/rebuild option. Money well, if reluctantly, spent in that regard.

To prove a point we reverted to a known working TB v2.3 build (the one we're migrating off) and all worked fine PRI-wise, so definitely not hardware.

Once I had a PiaF backup off-box, and the TB test above worked as it did, we rebuilt using the same options (PiaF v1.3 32-bit Asterisk 1.4), applied the restore, and it all just fired up as we needed it to. A beer-worthy event !

Anyway, once we had outbound calling again we had a subsequent weirdness happen with hangup detection, which lead to a version-specific libpri tweak+recompile as detailed here : http://pbxinaflash.com/community/threads/info-only-if-zap-calls-dont-hang-up.2674/?t=2674

Lesson #1 - if in doubt, rebuild at least once.

Kinda like Winduhz "if in doubt reboot" but at least here I don't have any expectation of having to do that every 4 days until I die ...

Lesson #2 - Backup/Restore still works

So, ultimately no great shakes in this process but I hope it helps someone later on.

Thanks for the bump euser4life. Realistically I prolly would have forgotten to go back to this.

 

[euser4life]

Glad you got it going. Sounds like it was a frustrating situation. Especially when you know you've had it working in the past and it should be working the same. Definitely makes you question your sanity

Thanks for sharing.