[Asterisk-Dev] Re: Inestability with H323

Sun Apr 18 19:24:26 MST 2004

Have you looked at the asterisk-oh323 driver?  That doesn't use
Asterisk's RTP stack.

On Sun, 2004-04-18 at 20:38, Derek Smithies wrote:
> Adam,
> 
> > Should we try and make the required changes to chan_h323 so it can 
> > compile with the latest open.H323. Derek, you're the OpenH.323 pro, 
> > fancy giving it a spin. Although it may not fix this bug, there's a few 
> > other bugs it will.. specifically the known exploits in the old openh323.
> 
> I am happy to fix the h.323 side of asterisk.
> Here are the problems I would face, should I take this job on.
> 
> a)the lack of response from others on this list who have been involved
>   with chan_h323.c   There is knowledge there that I want to tap into.
>   To be honest, I am not keen to "reinvent the wheel". I want to learn 
>   from others, and learn what trials/tribulations they had.
> 
> b)I have an architecture reorganisation in my mind that will mean 
>   chan_h323.c receives/sends encoded audio frames from/to the h323 stack
>   Currently, there appears to be the asterisk does all rtp handling. This
>    is not necessary, and is unwise in my view. I just need some clues as
>    to how to "connect" the encoded audio frames to asterisk.
> 
> c)upgrade of the chan_h323 to work with current  openh323 code.
>   Add sentences to the readme.
>    cvs co -r tagname pwlib
>    cvs co -r tagname openh323
>  
> d)you don't need to worry about exploits in the openh323 code.
>   iax2 has many more exploits than openh323.
> 
> Derek.
> =====================================================================
> On Mon, 19 Apr 2004, Adam Hart wrote:
> 
> > Derek Smithies wrote:
> > 
> > >Peter Nixon wrote:
> > >
> > >  
> > >
> > >>[*] Anyone who thinks otherwise is welcome to setup a copy of Asterisk running 
> > >>chan_h323 and tell me the IP. I guarantee that I can crash it for you in 
> > >>under 5 minutes, using only inbound H323 calls
> > >>    
> > >>
> > >
> > >
> > >Yep, and if you don't want Peter crashing your *for you, I will do it.
> > >
> > >I have detailed in a previous email how to do this. With just one external 
> > >computer, generating one call at a time, guaranteed failure.
> > >
> > >Peter Nixon also wrote:
> > >  
> > >
> > >>Until chan_h323 is fixed I would not rely on it for commercial
> > >>operation.
> > >>    
> > >>
> > >I will not rely on * + h323 for commercial operation either.
> > >
> > >  
> > >
> > Should we try and make the required changes to chan_h323 so it can 
> > compile with the latest open.H323. Derek, you're the OpenH.323 pro, 
> > fancy giving it a spin. Although it may not fix this bug, there's a few 
> > other bugs it will.. specifically the known exploits in the old openh323.
> > 
> > 
> > >Derek.
> > >
> > >=====================================================================
> > >On Mon, 19 Apr 2004, Peter Nixon wrote:
> > >
> > >  
> > >
> > >>On Sunday 18 April 2004 11:16, sales at minixel.com wrote:
> > >>    
> > >>
> > >>>I get this error every time when I reach 33 open calls
> > >>>using H323:
> > >>>-- Executing Dial("SIP/16468370490-0751",
> > >>>"H323/16468370490 at 192.168.85.100/16468370490") in new
> > >>>stack
> > >>>    -- Called 16468370490 at 192.168.85.100
> > >>>  0:14.917         H225 Answer:41802ba8
> > >>>assert.cxx(105)   PWLib   Assertion fail: Invalid array
> > >>>index, file /root/pwlib/include/ptlib/contain.inl, line
> > >>>423, Error=22
> > >>>
> > >>>
> > >>>I checked the header file referenced on the error and
> > >>>there is an object array being queried, and the object
> > >>>is not there. The question is: is this something
> > >>>fixable on our code, or in the pwlib libraries, or
> > >>>both? Any suggestion? I need to provide Sip-to-H323
> > >>>services. This feature is what makes my client be
> > >>>interested. In order to test this problem I generated a
> > >>>simple loop within Asterisk, sending the callback into
> > >>>Asterisk via H323:
> > >>>exten => _X.,1,ChanIsAvail(Sip/${EXTEN})
> > >>>exten => _X.,2,Dial(Sip/${EXTEN})
> > >>>exten =>
> > >>>_X.,102,Dial(H323/${EXTEN}@192.168.85.100/${CALLERIDNUM})
> > >>>
> > >>>By the way I have captured around 20 Core Dumps. What
> > >>>is exactly what I am supposed to look at? Sorry for my
> > >>>lack of expertise, but I figured out the ODBC problem
> > >>>and fixed it. I am getting there.
> > >>>      
> > >>>
> > >>The chan_h323 is still horribly unstable for any type of inbound bulk calls. * 
> > >>It may or may not be stable for originating bulk h323 calls but I wouldn't 
> > >>bet on it.
> > >>
> > >>I believe Jeremy is working on the issue. Until chan_h323 is fixed I would not 
> > >>rely on it for commercial operation. If you need a SIP-H323 proxy there are 
> > >>plenty available other than asterisk. If you need the extra "PBX" features of 
> > >>Asterisk, then unfortunately you have to wait..
> > >>
> > >>[*] Anyone who thinks otherwise is welcome to setup a copy of Asterisk running 
> > >>chan_h323 and tell me the IP. I guarantee that I can crash it for you in 
> > >>under 5 minutes, using only inbound H323 calls
> > >>
> > >>
> > >>    
> > >>
> > >
> > >  
> > >
> > 
> > _______________________________________________
> > Asterisk-Dev mailing list
> > Asterisk-Dev at lists.digium.com
> > http://lists.digium.com/mailman/listinfo/asterisk-dev
> > To UNSUBSCRIBE or update options visit:
> >    http://lists.digium.com/mailman/listinfo/asterisk-dev
> > 
> > 
> > 
-- 
          Eric Wieling * BTEL Consulting * 504-899-1387 x2111
"In a related story, the IRS has recently ruled that the cost of Windows
upgrades can NOT be deducted as a gambling loss."