Hi,

Just adding some information, RFC3261 allows sending re-invite without SDP 
body. I don't think this is a good idea because the UA witch sends a re-invite 
with a new renegotiation should present its alternatives. I remmember that RFC 
says 'MAY', witch preciselly means it's possible but it's not a good idea. 


Section 14.1 UAC Behavior (page 86)

    (...)It is important to note that the full
    description of the session, not just the change, is sent. This
    supports stateless session processing in various elements, and
    supports failover and recovery capabilities. Of course, a UAC MAY
    send a re-INVITE with no session description, in which case the first
    reliable non-failure response to the re-INVITE will contain the offer
    (in this specification, that is a 2xx response)


Anyway,  about the re-invite renegotiation, also from RFC 3261:

14.2 UAS behavior (page 88)

    If the new session description is not acceptable, the UAS can reject
    it by returning a 488 (Not Acceptable Here) response for the re-
    INVITE. This response SHOULD include a Warning header field.

I've seen the signaling, and there is no problem with that. The INVITE goes 
with the SDP body and the 200 OK response confirms compability. In re-INVITE 
signaling the same thing happens. 
Then, reading the Johnson's description, if the PBX or the ISP doesn't support 
G.729, it should not accept any call at all. Not even for sending a new 
re-INVITE after. Although, some implemmentations actually does that. Agein, RFC 
3261 recommends waht I'm talking about:

    8.1.3 Processing Responses

    8.1.3.5 Processing 4xx Responses (page 45)

    If a 415 (Unsupported Media Type) response is received (Section
    21.4.13), the request contained media types not supported by the UAS.
    The UAC SHOULD retry sending the request, this time only using
    content with types listed in the Accept header field in the response,    
    with encodings listed in the Accept-Encoding header field in the
    response, and with languages listed in the Accept-Language in the
    response.

    21.4.13 415 Unsupported Media Type (page 187)
    The server is refusing to service the request because the message
    body of the request is in a format not supported by the server for
    the requested method. The server MUST return a list of acceptable
    formats using the Accept, Accept-Encoding, or Accept-Language header
    field, depending on the specific problem with the content. UAC
    processing of this response is described in Section 8.1.3.5.


Then, assuming that the PBX or the ISP doesn't support G.729, it should have 
one of following two actions:

1) If it is specified in initial INVITE SDP body, it should respond with 200 OK 
accepting the call only with codec G.711 in SDP body. 

2) If it is not specified in initial INVITE SDP body, it should respond the 
initial INVITE with 415 Unsupported Media Type with Accept header specifying 
codec G.711.  Then, the UAC shall try a new call using G.711.

 

Leonardo


________________________________
De: "Johnson, Michael A" <michael.a.john...@team.telstra.com>
Para: "Worley, Dale R (Dale)" <dwor...@avaya.com>; 
"sip-implementors@lists.cs.columbia.edu" 
<sip-implementors@lists.cs.columbia.edu>
Enviadas: Quinta-feira, 30 de Junho de 2011 20:45
Assunto: Re: [Sip-implementors] Re-Invite codec renegotiation.

Sorry Dale, 
I thought it was clear.
The PBX (Mitel) that sends the re-invite without SDP, that then accepts the 
offered G.729 codec from the ISP, despite not being able support the codec is 
the device I believe is as fault. The manufacturer however disagrees.

As per the comment
"Although it might look like we're not doing it correctly, we are. ACK is the 
proper msg to be sent to acknowledge a response (2xx-6xx) to an INVITE request 
as per RFC3261"

I believe they should reject the G.729 offer. This does not appear to be 
clearly defined in the RFC




-----Original Message-----
From: Worley, Dale R (Dale) [mailto:dwor...@avaya.com] 
Sent: Friday, 1 July 2011 12:51 AM
To: Johnson, Michael A; sip-implementors@lists.cs.columbia.edu
Subject: RE: Re-Invite codec renegotiation.

________________________________________
From: Johnson, Michael A [michael.a.john...@team.telstra.com]

Thanks Dale,
The problem I have is that I cannot find anything definitive in RFC3261 that 
explicitly states that the system cannot accept the 200OK despite the fact it 
has no capability to handle the codec. The best I see in the RFC is 'SHOULD'.
To get the code corrected, I really need a clearly documented breach in the way 
they handle the call. And while it is obvious that this is not correct 
behaviour, this is not what I get back from their design team:

"Although it might look like we're not doing it correctly, we are. ACK is the 
proper msg to be sent to acknowledge a response (2xx-6xx) to an INVITE request 
as per RFC3261"

They have implemented a dodgy work around, however I want a proper fix and was 
seeking assistance to find the exact RFC extract to prove they are in breach of 
the standards.
________________________________________

Please...  You still haven't clearly stated which device you think is doing 
something wrong, and exactly what you think it is doing wrong.  I can't afford 
to spend more time on this discussion until you make it clear *what the problem 
is*.

Dale

_______________________________________________
Sip-implementors mailing list
Sip-implementors@lists.cs.columbia.edu
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
_______________________________________________
Sip-implementors mailing list
Sip-implementors@lists.cs.columbia.edu
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to