Congestion (fast busy) problem

We have a call center setup and I have witnessed that every 1 out of 10 or so calls we will get the congestion (fast busy) signals on a Zap channel even before it is rung although the number can be dialed without a problem if you dial it again right after. The Zap channels comes back with “Channel 0/5, span 1 got hangup request”. It is almost as Asterisk is trying to grab a Zap channel that is grabbed by another caller at the same time or trying to grab a channel that has not been released yet.

Any ideas?

Here is the CLI output of one:

-- Executing Macro("SIP/158-b7a7", "outbound-setup") in new stack
-- Executing Dial("SIP/158-b7a7", "Zap/g1/numbertodial||tTW") in new stack
-- Requested transfer capability: 0x00 - SPEECH
-- Called g1/numbertodial
-- Zap/5-1 is proceeding passing it to SIP/158-b7a7
-- Channel 0/5, span 1 got hangup request
-- Zap/5-1 is busy
-- Hungup 'Zap/5-1'

== Everyone is busy/congested at this time (1:1/0/0)
– Executing Congestion(“SIP/158-b7a7”, “”) in new stack
== Spawn extension (international, 9numbertodial, 3) exited non-zero on ‘SIP/158-b7a7’

yeah, we had something like this happen.

our telco is delivering calls in order of ascending channel. that means they start at channel 1, and if that’s in use go to channel 2, and so on.

we were dialing out on our Zap trunks in ascending order also [Dial(Zap/g1/…] which was causing some collisions.

Changing the dial order to DEscending [Dial(Zap/G1/…] fixed the issue completely.

Not sure if this would fit your circumstances, but it might…

Yep it does, cannot believe I didn’t think of that. Thanks Wes!

glad to help - took me about a month to figure that one out, and we even opened a ticket with the telco…finally somebody there asked how we were routing calls.

sometimes it’s the simplest things… :smiley:

True sometimes we make the issue more difficult than it is, and the obvious always goes unnoticed :smile: