Hi, 

>i thought that dean meant with "dual registration" registering with two
>different registrars (e.g. two srv records for the same domain).

Not sure, perhaps. Hope that Dean will clarify what he meant and
whether I misunderstood his terminology.

>that case, your problem does not occur if proxies are also paired with
>the two registrars.

But even in that case I can not follow. Yes I see that is possible
without Outbound but on the other hand I fail to see Outbound providing
any added value for such a setup.

Or is the point exactly that :

- without Outbound you need multiple different registrars (having
different
  srv records for them) for such dual registration ?

- but with Outbound you could do similar thing with one single registrar
  while having multiple (or multihomed) proxies ?

Regards,

Erkki

>-----Original Message-----
>From: ext Juha Heinanen [mailto:[EMAIL PROTECTED] 
>Sent: 06.October.2008 08:51
>To: Koivusalo Erkki (Nokia-D/Helsinki)
>Cc: [EMAIL PROTECTED]; [email protected]
>Subject: Re: [Sip] Dual registration without Outbound
>
>[EMAIL PROTECTED] writes:
>
> > Problem with dual registrations without Outbound comes
> > with forking. When a device registers itself twice with
> > different Contact addresses, it will receive every request
> > destined to the AOR twice if the server implements parallel
> > forking or if the device rejects the first forked request.
>
>erkki,
>
>i thought that dean meant with "dual registration" registering with two
>different registrars (e.g. two srv records for the same domain).  in
>that case, your problem does not occur if proxies are also paired with
>the two registrars.
>
> > A long time ago I posted a long email suggesting Outbound
> > to be split to two drafts, addressing two different problems:
> > 
> > - One that specifies the keepalive behaviour (needed for NATs)
> > - One that specifies multiple registration behaviour
> >   (needed for increased reliability) 
>
>and third draft should be about connection re-use.  
>
>if iesg or whoever still exists whose task is to keep rfcs sane, it
>should definitely bounce outbound i-d back to the working group so that
>it can do the right thing rather that rushing outbound that only
>includes application specific solution to these generic problems.
>
>-- juha
>
>
_______________________________________________
Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [EMAIL PROTECTED] for questions on current sip
Use [EMAIL PROTECTED] for new developments on the application of sip

Reply via email to