In article <4291695B.6060303@empoweredcomms.com.au>,
Rod Bacon <rod.bacon@empoweredcomms.com.au> wrote:> I'm new to H.323 and I have noticed that there are two separate channel
> drivers for * available - the inbuilt one, and oh-323. I had trouble
> compiling oh-323 with the current cvs stable, so I tried the inbiult one
> (with specifiec recommended versions of openh323 and pwlib). It compiled
> cleanly but I am told that it is not recommended (unstable?).
>
> Can someone with first-hand * H.323 experience offer any meaningful
> advice as to which way I _must_ proceed? This is for a live, busy,
> deployed environment. H.323 will be used to connect to an upstream
> provider (possibly CISCO gear?).
I'm using stable, and first of all tried the built-in chan_h323. It
compiled ok, but when I tried to get two Asterisk boxes to talk to each
other using H.323, they could set up calls but not pass any audio.
So then I moved on to chan_oh323 v0.6.5 with the Janus patch 4 releases
of openh323 (1.13.5.3) and pwlib (1.6.6.3). This performs mostly OK,
but I do have an issue with intermittent audio flutter on the output
stream, which I think is due to spurious repeated packets. I'm still
trying to get to the bottom of that (see my other posts on this list).
If I can resolve that, then I would recommend chan_oh323.
There are two other H.323 implementations for Asterisk (ooh323 and
chan_woomera), but they both require CVS-HEAD instead of stable :-(
I'll try them out when HEAD stops being so much of a moving target.
At the very least, you should make sure you can get two Asterisk boxes
to talk to each other properly using H.323 before you try talking to the
upstream provider. Also, you will need to get your upstream provider to
disable VAD (Silence suppression) in the streams it sends you.
Hope this helps
Cheers
Tony
--
Tony Mountifield
Work: tony@softins.co.uk - http://www.softins.co.uk
Play: tony@mountifield.org - http://tony.mountifield.org