The from and to tags also need to be considered.

If an INVITE request received provisional (18x) dialog establishing
responses from multiple UAS's and reliable provisional responses (RFC 3262)
were being used. The resulting PRACKs would have the same call-id & seq# but
different to-tags. If only the call-id & seq# were used to generate a branch
id, both PRACK requests would have the same branch-id.

Also, in the case of a forking proxy, it will need to add something to make
the branch parameter unique for each branch of the fork.

cheers,
(-:bob

Robert F. Penfield
Chief Software Architect
Acme Packet, Inc.
130 New Boston Street
Woburn, MA 01801
[EMAIL PROTECTED]

----- Original Message ----- 
From: "Brett Tate" <[EMAIL PROTECTED]>
To: "'Sudharsana Krishnan'" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Monday, September 15, 2003 5:24 PM
Subject: RE: [Sip-implementors] Question on creation of branch
parameteratthe Transaction Layer


> > I assume you are looking at this as a UAC
> > and not a proxy since this approach makes
> > no sense for a proxy at all.
> >
> > You'll be ok until the first time you have
> > to send a request to a second destination
> > while working through the algorithm in RFC3263.
> > At that point, you'll be broken.
>
> Another issue is converting definition of a call-id
> (words and '@') into the reduced syntax of a
> via branch (token).
>
> Another potential issue is associated with the
> interpretation of the "globally unique" statement
> in rfc3261 section 8.1.1.7.  If this only
> applies in relation to source address, there is
> no problem.  Otherwise there will potentially be
> a problem if the other UA uses the same algorithm
> and starts with approximately the same cseq.
>
> > On Mon, 2003-09-15 at 09:31, Sudharsana Krishnan wrote:
> > > All
> > >
> > > I had a question related to creation of the branch
> > > parameter for outgoing requests.
> > >
> > > I want to use the CSeq number appended to the
> > > call-id in the sip request as
> > > the branch parameter.
> > > (after the magic cookie)
> > >
> > > Since Call-ID will be globally and spatially unique,
> > > what I am doing I think
> > > will ensure uniqueness for branch creations.
> > >
> > > Does somebody out there forsee an issue?
>
> _______________________________________________
> Sip-implementors mailing list
> [EMAIL PROTECTED]
> http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
>

_______________________________________________
Sip-implementors mailing list
[EMAIL PROTECTED]
http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors

Reply via email to