asterisk users - Aug 2007 - Asterisk 1.2 + Zaptel 1.4 + HPEC = Crash?

Hello,

Has anyone tried the combination of asterisk 1.2.24, zaptel 1.4.5 and 
HPEC 9.00.003?

In particular, with a hardware configuration similar to:

Module 0: Installed -- AUTO FXO (FCC mode)
Module 1: Installed -- AUTO FXO (FCC mode)
Module 2: Installed -- AUTO FXO (FCC mode)
Module 3: Not installed
Found a Wildcard TDM: Wildcard TDM400P REV I (3 modules)

I have two fully independent systems (both production, so I can't do 
further testing unfortunately) that crash anywhere between an hour and a 
day after booting under a minimal load.  If HPEC is disabled, the 
problem is gone (but really bad echo).  If I use zaptel 1.2.20.1, the 
problem is gone.

The result is a kernel panic followed by an automatic reboot.  Nothing 
is written to log files so I cannot provide any debug information.  As 
mentioned this has happened on multiple production machines and I do not 
have any other "wctdm" cards to test with.

I would be curious to hear if anyone else noticed the same problem or if 
they have it working.  What are the common denominators?

Thanks,
Trevor

-- 
Does your Canadian VoIP service need CRTC-compliant 9-1-1 services?  Please
visit http://www.digitalcon.ca/voip9-1-1/ to find out more!

Trevor Peirce wrote:
> Hello,
> 
> Has anyone tried the combination of asterisk 1.2.24, zaptel 1.4.5 and 
> HPEC 9.00.003?
First of all... Why are you using zaptel 1.4.5 with asterisk 1.2?  That 
is a red flag in itself.
> 
> In particular, with a hardware configuration similar to:
> 
> Module 0: Installed -- AUTO FXO (FCC mode)
> Module 1: Installed -- AUTO FXO (FCC mode)
> Module 2: Installed -- AUTO FXO (FCC mode)
> Module 3: Not installed
> Found a Wildcard TDM: Wildcard TDM400P REV I (3 modules)
> 
> I have two fully independent systems (both production, so I can't do 
> further testing unfortunately) that crash anywhere between an hour and a 
> day after booting under a minimal load.  If HPEC is disabled, the 
> problem is gone (but really bad echo).  If I use zaptel 1.2.20.1, the 
> problem is gone.
> 
> The result is a kernel panic followed by an automatic reboot.  Nothing 
> is written to log files so I cannot provide any debug information.  As 
> mentioned this has happened on multiple production machines and I do not 
> have any other "wctdm" cards to test with.
> 
> I would be curious to hear if anyone else noticed the same problem or if 
> they have it working.  What are the common denominators?
> 
> Thanks,
> Trevor
> 

-- 
Matthew Fredrickson
Software/Firmware Engineer
Digium, Inc.

Trevor Peirce wrote:
> Has anyone tried the combination of asterisk 1.2.24, zaptel 1.4.5 and 
> HPEC 9.00.003?
>   
[snip]
> The result is a kernel panic followed by an automatic reboot.  Nothing 
> is written to log files so I cannot provide any debug information.  As 
> mentioned this has happened on multiple production machines and I do not 
> have any other "wctdm" cards to test with.
>   
Although I can't say for certain that it's not Asterisk 1.2 vs. Zaptel 
1.4 causing your issues, I've had crashes with HPEC on Zaptel 1.2 as 
well, and finally tracked these down to the HPEC binary overwriting bits 
of memory it shouldn't be touching.

It may be the case that it's doing this for you on Zaptel 1.2, but due 
to code differences, isn't overwriting anything critical (for me, the 
crashes only started once we moved from kernel 2.6.15.7 to 2.6.20.x - 
the memory layout changed, HPEC would quickly corrupt the free list, and 
things would go boom upon the second deallocation of an HPEC EC).

I've been working with Digium to get this found and fixed, so with luck 
there'll be an updated HPEC sometime reasonably soon.

Cheers,
    Nic.

-- 
Nic Bellamy,
Head Of Engineering, Vadacom Ltd - http://www.vadacom.co.nz/