[asterisk-users] IAX2 was Re: Problems with 2 Asterisk servers on same LAN
Tim Panton
thp at westhawk.co.uk
Mon Sep 8 08:37:11 CDT 2008
On 8 Sep 2008, at 13:12, Steve Totaro wrote:
> On Sun, Sep 7, 2008 at 9:57 AM, Michiel van Baak
> <michiel at vanbaak.info> wrote:
>> On 08:24, Sun 07 Sep 08, Steve Totaro wrote:
>>>
> Maybe the problem is that IAX2 is not as set in stone as the RFCs for
> SIP? Who is to say it is or isn't "compliant" to the guidelines?
>
Which of the 100+ SIP/STUN/RTP RFCs with all their optional
appendixes are we talking about ?
Standardization really _isn't_ SIP's strong point.
(Wide adoption is )
Some of us (Ed Guy in particular) are making an effort to get
the IAX2 RFC into shape.
The draft is pretty usable now.
http://www.ietf.org/internet-drafts/draft-guy-iax-04.txt
It has the huge advantage that
as it is 108 pages, you at least stand a chance of
knowing what should happen :-)
It is true that the implementation in asterisk has changes
and there have been the odd snafu's but that's been true
of all the features.
>>
>>>
>>> I think I would spend a day or two getting SIP working properly,
>>> now,
>>> rather than spending days trying to figure out audio issues and
>>> having
>>> to revisit and get SIP working properly in the future. After people
>>> are actually relying on the system and already have a bad
>>> experience/opinion associated with the "New Phone System".
>>
>> I think they should go with the Tech their DID provider prefers.
>> That way you will get the best support from them if something goes
>> wrong.
>>
>>
>
> I sort of agree although I doubt I would use a provider that pushed
> IAX2 except as an option. It has also been my experience that when
> IAX2 does not work well, the providers do not offer much support
> beyond the initial configs, failing that, they suggest using SIP as
> well.
Like I said to the OP, if your provider supports it, then it will
probably be
simpler to implement at your end in this specific case (multiple
asterisks behind a single NAT supporting incoming calls).
Tim.
>
More information about the asterisk-users
mailing list