[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Sipping] Is SDP in an unreliable response "the answer" ???



Hi,

>Thanks for identifying this. I should have remembered.
>
>So that is a case we don't have to cope with.

>From my experience the 200 OK will NOT be delayed, and that might be one reason why people include the SDP answer in the 200 OK, even if it has been sent reliably earlier - at least if the response in which the SDP answer was sent has not been acknowledged when the 200 OK is to be sent.

I am not sure we would need to describe this, since the behavior breaks 3262. However, it does happen, and it is a reason why people include SDP in 200 OK, so maybe a Note could be useful? 

Regards,

Christer


Anders Kristensen wrote:
>
>
> On 4/20/2010 7:04 AM, Paul Kyzivat wrote:
>>
>>
>> DRAGE, Keith (Keith) wrote:
>>> Do remember that there are certain cases where the UAC will not ignore
>>> it. These are the cases where the message containing original answer
>>> has not yet arrived or been lost, and the protocol has not yet
>>> recovered from that.
>>
>> Hmm. Can you say more?
>>
>> Are you thinking of a case where the answer has been sent in a reliable
>> provisional, but the PRACK has not yet been received? And then that
>> *some* SDP is included in a subsequent unreliable provisional, or 2xx?
>>
>> That is an interesting case. The one that would be least wierd is:
>>
>> UAC UAS
>> | INVITE (SDP1) |
>> |----------------->|
>> | 1xx REL (SDP2) |
>> | X--------------|
>> | 2xx (no SDP) |
>> |<-----------------|
>
> This is not a valid case. It's explicitly ruled out by 3262:
>
>    If the UAS had placed a session description in any reliable
>    provisional response that is unacknowledged when the INVITE is
>    accepted, the UAS MUST delay sending the 2xx until the provisional
>    response is acknowledged.  Otherwise, the reliability of the 1xx
>    cannot be guaranteed, and reliability is needed for proper operation
>    of the offer/answer exchange.
>
> Thanks,
> Anders
>
>>
>> I think that implies a different best practice:
>>
>> If the UAS has sent an answer in a reliable provisional, if it sends a
>> 2xx before receiving a prack of the answer, it should include the answer
>> in the 2xx as well.
>>
>> (If the UAS intends to initiate another o/a it MUST await the prack, so
>> if the 1xx is lost it will be retransmitted.)
>>
>> Thanks,
>> Paul
>>
>>
>>
>>
>> | |
>> |<-----------------|
>> | |
>> | |
>> |----------------->|
>> | |
>> |<-----------------|
>> | |
>> | |
>> |----------------->|
>> | |
>> |<-----------------|
>> | |
>> | |
>> |----------------->|
>> | |
>> |<-----------------|
>> | |
>> | |
>> |----------------->|
>> | |
>> |<-----------------|
>> | |
>>
>>> regards
>>>
>>> Keith
>>>
>>>> -----Original Message-----
>>>> From: sipping-bounces@xxxxxxxx [mailto:sipping-bounces@xxxxxxxx] On
>>>> Behalf Of Christer Holmberg
>>>> Sent: Tuesday, April 20, 2010 1:47 PM
>>>> To: OKUMURA Shinji; sipping@xxxxxxxx
>>>> Subject: Re: [Sipping] Is SDP in an unreliable response "the answer"
>>>> ???
>>>>
>>>>
>>>> Hi,
>>>>>>> Before sending an answer,
>>>>>>> - An UAS MAY send unreliable provisional responses with a SDP.
>>>>>>> - And the SDP MUST be identical to an answer SDP.
>>>>>>>
>>>>>>> After sending an answer,
>>>>>>> - The UAS should not insert a SDP in any response.
>>>>>>>
>>>>>>> Is this OK?
>>>>>> That text still doesn't say what an SDP inserted after sending the
>>>>>> answer means, only that it should not be sent.
>>>>> The SDP means nothing. it is neither an offer nor an answer.
>>>> Exactly. In my opinion that is what is important - not whether the
>>>> UAS inserts SDP or not.
>>>>
>>>>>> I still don't see why we need to make a separation about SDP sent
>>>>>> before and after the answer, because in both cases the SDP must be
>>>>>> identical to the answer.
>>>>> 1. RFC3261 says that UAS MAY send it before the answer and
>>>>> doesn't say nothing after the answer.
>>>> That is one reason why we are writing the draft - to clarify things
>>>> which may not be clear in the specs.
>>>>
>>>>> 2. The SDP MUST be ignored by UAC. it is meaningless.
>>>> I agree, and that is what we must be clear about. Because, as we
>>>> know, some people want to send a NEW offer (or updated answer) in a
>>>> subsequent response, and that is not allowed.
>>>>
>>>>
>>>>> 3. if another o/a exchange is occured (using UPDATE or
>>>> PRACK), it is not even a confirmation.
>>>>> And, again, I know there are many implementations that send
>>>> a copy of
>>>>> the SDP after the SDP answer has been sent, so instead of
>>>> saying that
>>>>> it should not be done I think it is much more important to
>>>> say that, if
>>>>> it is done, it must be identical to the SDP answer. In other
>>>> words, to
>>>>> make it clear that the UAS can not send a NEW offer (or
>>>> updated answer)
>>>>> in a subsequent response after the SDP answer has been sent.
>>>>>
>>>>> Since UAC MUST ignored it, there is no problem on a interworking.
>>>>> Why must it be identical to the SDP answer?
>>>> Well, if you look at it that way, fine. But, then the important thing
>>>> is that the UAC must ignore it - not that the UAS should not send it.
>>>>
>>>> Regards,
>>>>
>>>> Christer
>>>>
>>>>
>>>>
>>>>
>>>>>> Regards,
>>>>>>
>>>>>> Christer
>>>>>>
>>>>>> Hans Erik van Elburg <ietf.hanserik@xxxxxxxxx> Mon, 19 Apr 2010
>>>>>> 13:55:41 +0200
>>>>>>>> - An UAS MAY insert a SDP body that is identical to the
>>>>> SDP answer,
>>>>>>>> in an unreliable provisional response before the SDP
>>>> answer has
>>>>>>>> been sent.
>>>>>>>>
>>>>>>>> - The UAS MUST NOT insert a SDP body that is not
>>>>> identical to the
>>>>>>>> SDP answer, in an unreliable provisional response
>>>> before the SDP
>>>>>>>> answer has been sent.
>>>>>>>>
>>>>>>> This is terribly confusing. Very probabe that noone will
>>>>> get it right.
>>>>>>> Triple negation. And talking about sending and answer before the
>>>>>>> answer has been sent. ???
>>>>>>>
>>>>>>>
>>>>>>>> - The UAS MUST NOT insert a SDP body in any response
>>>>> after the SDP
>>>>>>>> answer has been sent.
>>>>>>>>
>>>>>>> This means that you can't send it again after you've send
>>>> it in an
>>>>>>> unreliable provisional response. Do you want tto say that?
>>>>>>>
>>>>>>> /Hans Erik van Elburg
>>>> _______________________________________________
>>>> Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping
>>>> This list is for NEW development of the application of SIP Use
>>>> sip-implementors@xxxxxxxxxxxxxxx for questions on current sip Use
>>>> sip@xxxxxxxx for new developments of core SIP
>>>>
>>> _______________________________________________
>>> Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping
>>> This list is for NEW development of the application of SIP
>>> Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
>>> Use sip@xxxxxxxx for new developments of core SIP
>>>
>> _______________________________________________
>> Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping
>> This list is for NEW development of the application of SIP
>> Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
>> Use sip@xxxxxxxx for new developments of core SIP
>>
> _______________________________________________
> Sipping mailing list  https://www.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
> Use sip@xxxxxxxx for new developments of core SIP
>
_______________________________________________
Sipping mailing list  https://www.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
Use sip@xxxxxxxx for new developments of core SIP
_______________________________________________
Sipping mailing list  https://www.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
Use sip@xxxxxxxx for new developments of core SIP

[IETF Announce]     [IETF Discussion]     [Linux SCSI]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Photo]     [Yosemite News]    [Yosemite Photos]    [Free Online Dating]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Big List of Linux Books]

Add to Google Powered by Linux