Hi Atul,

 <http://bugs.sipit.net/show_bug.cgi?id=648>see inline..
On 10/5/06, Atul kumar Jha <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
>
>
> I have some doubts regarding the Loop detection in case of Stateful
> Proxy.
>
>
>
> Q1. Should we include the top via branch ID in the branch ID calculation
> for Loop detection?


<amar>
Go thru the draft-ietf-sip-fork-loop-fix-03.txt. Basically you shouldn't
include the via as part of calculating the
branch parameter. As this may seem to a be a retransmition of message in the
other proxy who is not doing Loop detection.

In stead of taking the via any other values used by the location service
logic while processing the request.
 </amar>

Q2. And in the case of spiral should we suppress the Record Route or not
> as it might lead to extra processing load for the Proxy?


<amar>
Thats true. Even this doesn't provide any value add. Just decreases the
proxy performance. So record-routing can be supressed
in case of the spirals. But at the same time you always need to ensure that
if the RRH is not added when request arrived for
the first time at the proxy, and the same need to be added at the time of
spiral request (iff proxy does record routing at that
time). This could be the boudary condition if there is a dynamic
configuration property at the proxy end.
</amar>
>
> Thanking you in advance.
>
> Regards,
>
> Atul
>
>
>
> _______________________________________________
> Sip-implementors mailing list
> [email protected]
> https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>



-- 
Rgds,
Amar
Mobile: +919886395894
The greatest enemy of best is "good." If you're willing to accept "good"
you'll never be the "Best."
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to