On 04/23/2011 02:43 PM, Iñaki Baz Castillo wrote:
> 2011/4/22 Vijay K. Gurbani<[email protected]>:
>> Unlike rfc2543 where a received 6xx was immediately forwarded, in
>> rfc3261 this is not the case. A received 6xx response is not
>> quarantined until all other branches have generated a final
>> response.
>
> I don't think so. From RFC 3261:
See my follow-up response --- I meant that a received
6xx response *is* quarantined until other branches
finish. So either 6xx is the best response if all
branches had a URI rooted in the invalid domain, or
some other branch generated a better response.
> This is, from the UAC point of view, if no branch replies 200 and
> any branch replies 6XX then the UAC will receive such 6XX.
The UAC does not know that a downstream proxy forked, and
neither should it. Regardless of whether the proxy used
sequential or parallel forking, when the proxy sees a
6xx on one branch, it will send CANCELs on the remaining
n-1 branches and return the best response. If any of
the other branches had a different URI than the invalid
one you mention, then this becomes the best response.
If all branches had the URI rooted in the invalid
domain, then the best response is still the 6xx-class
one.
So I really do not understand the problem. Things
work as expected.
What are we missing here?
- vijay
--
Vijay K. Gurbani, Bell Laboratories, Alcatel-Lucent
1960 Lucent Lane, Rm. 9C-533, Naperville, Illinois 60566 (USA)
Email: vkg@{bell-labs.com,acm.org} / [email protected]
Web: http://ect.bell-labs.com/who/vkg/
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
