ok we are agree how proxy should handle an incoming call to UAs which are
registered with
same pin, passwd, uri but different contact address. we use what location
service record for a registration, call-id, Cseq, uri and contact address to
proven the behviour of proxy.

but i could not see any clear comment in the 3261 which claries this issue.

could you please tell me where i should see in the 3261. I lookad at the all
seciton 10...

Rgrds,

erol



On 2/25/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>
>   From: "Bob Penfield" <[EMAIL PROTECTED]>
>
>   This is not correct. If the two contact addresses are different, there
> will
>   still be two bindings for the address-of-record. The fact that the two
>   REGISTER messages have the same Call-ID does not cause the second
> REGISTER
>   to "replace" the binding of the first unless it is the same address.
>
> Ugh, yes, that's my mistake.  The rules are in section 10.3, section
> 7, though they're not clearly presented.
>
> The Call-Id and CSeq values (which define the pseudo-dialogs) are used
> to determine which REGISTERs can revise the bindings created by which
> REGISTERs.  The processing is done on a per-contact-URI basis, if two
> REGISTERs that contain the same contact have the same Call-Id, the one
> with the largest CSeq wins, but if the two REGISTERs have different
> Call-Ids, the one that arrives last wins.
>
> So if one pseudo-dialog contains updates for two contacts, each
> contact is updated independently of the other, as if the REGISTERs not
> mentioning that contact don't exist.
>
> In the perverse case of two pseudo-dialogs containing updates for one
> contact, the likely result is that the last REGISTER to arrive at the
> registrar will take precedence, rather than the registrar detecting
> out-of-order REGISTERs using the CSeq value.
>
> Dale
> _______________________________________________
> Sip-implementors mailing list
> [email protected]
> https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>



-- 
Erol TuraƧ
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to