[Serusers] r-uri for ACK

[Richard]

Hi Jiri,

If I have record-route, ACK to a 200 OK should be forwarded loose route. But
for an ACK to negative reply, e.g. 487, it uses the original r-uri. Because
ACK to negative reply is hop-by-hop basis, ser should absorb the ACK and
won't forward it further, right? In ser.cfg, should we just drop it in case
ser doesn't absorb it? Downstream won't understand the ACK anyway and might
be screw up any existing call...

Richard


> -----Original Message-----
> From: Jiri Kuthan [mailto:jiri at iptel.org]
> Sent: Thursday, October 14, 2004 12:33 AM
> To: Richard; serusers at lists.iptel.org
> Subject: Re: [Serusers] r-uri for ACK
> 
> It depends. ACK for negative replies must have identical URIs as INVITEs.
> Otherwise, it is dicated by record-routing. Loose routers (i.e., those
> that
> implement RFC3261 as opposed to the obsoleted RFC2543) put peer's contact
> in there, which is ideal. Strict routers put record-routing information
> there. This alternative is valid in terms of an obsoleted spec.
> 
> -jiri
> 
> At 11:02 AM 10/14/2004, Richard wrote:
> 
> >Hi,
> >
> >I have a basic sip question. Whats the correct r-uri for ACK? I use
> stateful forwarding, so all SIP messages pass through ser. I have seen two
> types of UA. Some use the contact field of 200 OK response as the r-uri
> and other use the original r-uri for INVITE.
> >
> >Is it a SIP violation to use the original r-uri of INVITE? The problem
> here is that if there is a parallel forking for the INVITE, it might be
> sent to places other than the real callee.
> >
> >Any comment?
> >
> >Thanks,
> >
> >Richard
> >_______________________________________________
> >Serusers mailing list
> >serusers at lists.iptel.org
> >http://lists.iptel.org/mailman/listinfo/serusers
> 
> --
> Jiri Kuthan            http://iptel.org/~jiri/