Re: [hylafax-users] [?? Probable Spam] Re: Wedged Iaxmodems

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Maciej Żwirski wrote:
> > Can you maybe take a look at this particular time in the iaxmodem 
> > log?  (/var/log/iaxmodem/)  See if there are any clues there.  
> > Otherwise we'll need to enable debug logging for IAX2, and that gets 
> > very verbose.
> You were right! There was one more call afterwards, although it wasn't 
> "signed" with any device id, that's why I've missed it. Both Iaxmodem 
> and Asterisk logs don't show anything interesting (except that I was 
> able to trace the attached Hylafax log with them). It seems like the 
> transmissino I've sent you had some impact on the next transmission... 
> As for the human voice in the recordings... I've noticed that a lot of 
> people want to talk to our fax :) Some even sent in complains that the 
> fax is being rude and doesn't answer...
>
> gru 05 13:01:31.25: [32424]: SESSION BEGIN 000024765 XXXXXXX
> gru 05 13:01:31.25: [32424]: HylaFAX (tm) Version 4.3.1
> gru 05 13:01:31.25: [32424]: CallID: "YYYYYYY" "<NONE>" "<NONE>" 
> "ZZZZZZZZZ"
> gru 05 13:01:31.25: [32424]: MODEM set XON/XOFF/FLUSH: input ignored, 
> output disabled
> gru 05 13:01:31.25: [32424]: <-- [4:ATA\r]
> gru 05 13:01:31.29: [32424]: --> [0:]
> gru 05 13:01:31.29: [32424]: ANSWER: Unknown problem
> gru 05 13:01:31.29: [32424]: SESSION END 

And since I have a recording of this short-lived call already (and 
replaying that recording doesn't reproduce the crash) we know that it's 
not in the DSP.

So... the error is either to do with IAX2 handling, or in some kind of 
more general error (like a memory leak or something) that seems specific 
to your situation.

In iaxmodem.c you'll see this:

   /*
    * We disable debugging because the screen writes can slow things down.
    */
   iax_disable_debug();
//    iax_enable_debug();

You'll need to comment-out the "disable" line and uncomment the "enable" 
line, rebuild, reinstall, and restart.  Then you have to pay attention 
to the output... that should then be logged, but the logs will grow 
fast.  I'll need to see what happens in that logging when the crash 
occurs.  It's possible that the non-daemon output without the debugging 
will reveal the problem without the verbosity of the "debug" output... 
so if you can find that from this past error it would be nice to see.

Check your Asterisk logs.  Was there, by chance, another call being sent 
to the same iaxmodem almost simultaneous with this one?  So, were there 
two calls being sent to the same iaxmodem simultaneously here?

Thanks,

Lee.


____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*