[Asterisk-Users] Initial Connection Hangup (T100P) and Ringing Failure

[Matthew Farley]

I ran into a couple of odd problems with a new Asterisk box I have set
up... I have a T100P connected to a channelized T-1 (24ch) to the PSTN.
The T-1 is set up properly so far as I know and seems to work perfectly
with one exception:

When a call comes in, asterisk starts a 'simple switch' for it, goes
through my 's' extension progression until it gets to the first Playback
directive, then immediately detects that the Zap channel exited
non-zero, then hangs it up. This, however, does not hang up the call
from the caller's perspective. Instead, asterisk immediately starts a
'simple switch' on the same channel, and handles the call without a
hitch the second time around. From the caller's perspective, there is
just an abnormally long delay once the call is established before the
beginning of the asterisk prompts. Example console output is listed
below:

(Call comes in from PSTN)
   -- Starting simple switch on 'Zap/6-1'
   -- Executing NoOp("Zap/6-1", "") in new stack
   -- Executing Answer("Zap/6-1", "") in new stack
   -- Executing DigitTimeout("Zap/6-1", "5") in new stack
   -- Set Digit Timeout to 5
   -- Executing ResponseTimeout("Zap/6-1", "10") in new stack
   -- Set Response Timeout to 10
   -- Executing Background("Zap/6-1", "wconf-welcome") in new stack
   -- Playing wconf-welcome
  == Spawn extension (wstconf, s, 5) exited non-zero on 'Zap/6-1'
   -- Starting simple switch on 'Zap/6-1'
   -- Executing NoOp("Zap/6-1", "") in new stack
   -- Executing Answer("Zap/6-1", "") in new stack
   -- Executing DigitTimeout("Zap/6-1", "5") in new stack
   -- Set Digit Timeout to 5
   -- Executing ResponseTimeout("Zap/6-1", "10") in new stack
   -- Set Response Timeout to 10
   -- Executing Background("Zap/6-1", "wconf-welcome") in new stack
   -- Playing wconf-welcome

(Call continues normally from this point on, and never hangs up on the
PSTN caller at any point during this process)


Possibly related to this behaviour is the fact that if a caller is on a
Zap channel and hits an extension which contains the 'Ringing'
directive, asterisk segfaults. This does not happen if an extension
containing a 'Ringing' directive is called from an IAX or SIP phone. 

Has anyone ever seen these issues before? I'm at a bit of a loss myself,
and would appreciate any suggestions that you might have to offer. I'm
fairly sure that the T-1 itself is fine, as it works properly when
connected to another asterisk box with the same zaptel.conf and
zapata.conf settings.

Thanks!
-- 
Matthew Farley <asterisk at wheatstate.net>