[asterisk-users] Best Current Release for Long Term Use

[Tzafrir Cohen]

On Fri, May 29, 2009 at 09:24:02AM -0500, Danny Nicholas wrote:
> I beg to differ with you Jared.  Since I don't have your email, I'll post
> this here.  Create this call file.
> Channel: DAHDI/g1/5551212
> CallerID: SIP/104
> MaxRetries: 1
> WaitTime: 60
> retryTime: 5
> Application: playback
> Data: /var/lib/asterisk/sounds/you-sound-cute
> 
> On my "test" machine using a TDM400P, the playback would occur in a manner
> such that the user would only get 50-70 percent of the message.  
> 
> Changing 5551212 to a cell phone number would create another 5 to 10 percent
> loss of message.
> 
> Changing DAHDI/g1/5551212 to a local SIP extension would deliver the message
> in its entirety.
> 
> On my "live" machine (TDM410P), the DAHDI call actually waits for a
> connection, then plays the message after a 1/2 second pause, so I STAND
> CORRECTED, AT LEAST SORT OF.  The bug is reproducible on the TDM400P but not
> TDM410P.  
> 
> Here is dahdi_cfg from test box
> DAHDI Tools Version - 2.1.0.2
> 
> DAHDI Version: 2.1.0.3
> Echo Canceller(s):
> Configuration

Is this a regression from the past (some version of Zaptel)?

This is not merely a matter of casting the blame. If it is a regression,
it should be interesting to trace where is the exact point of
regression, which should help fix the problem.

It is aparantly a bug in the wctdm driver (TDM410P uses wctdm24xxp). Do
you have a "good" and a "bad" sample to compare alongside? It should be
interesting.

A good start would be to enable full debugging at Asterisk. This should
include all events sent from Zaptel/DAHDI to chan_zap / chan_dahdi.
Comparing the "good" and the "bad" should give a useful insight.

-- 
               Tzafrir Cohen
icq#16849755              jabber:tzafrir.cohen at xorcom.com
+972-50-7952406           mailto:tzafrir.cohen at xorcom.com
http://www.xorcom.com  iax:guest at local.xorcom.com/tzafrir