Hi Tang Xi,

In case of INVITE and re-INVITE UAC (UE) can send CANCEL message for INVITE or 
re-INVITE
but CANCEL will have effect as long as UAS has not already given final response 
(2XX).

More description:
CANCEL has no effect on a request to which a UAS has
already given a final response. Because of this, it is most useful
to CANCEL requests to which it can take a server long time to
respond.  For this reason, CANCEL is best for INVITE requests, which
can take a long time to generate a response.  In that usage, a UAS
that receives a CANCEL request for an INVITE, but has not yet sent a
final response, would "stop ringing", and then respond to the INVITE
with a specific error response (a 487).


Since requests other than INVITE are responded to immediately,
sending a CANCEL for a non-INVITE request would always create a
race condition.

See section Section 14 and Section 12 of RFC 3261 for more details.

Regards
Manmohan Singh Bisht    


On Wed, 22 Nov 2006 [EMAIL PROTECTED] wrote :
>Send Sip-implementors mailing list submissions to
>       [email protected]
>
>To subscribe or unsubscribe via the World Wide Web, visit
>       https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>or, via email, send a message with subject or body 'help' to
>       [EMAIL PROTECTED]
>
>You can reach the person managing the list at
>       [EMAIL PROTECTED]
>
>When replying, please edit your Subject line so it is more specific
>than "Re: Contents of Sip-implementors digest..."
>
>
>Today's Topics:
>
>    1. Re: Maximum size of Header filed Value (Attila Sipos)
>    2. Query cancel re-INVITE (Tang Xi)
>    3. Re: Re-Invite Cseq ([EMAIL PROTECTED])
>    4. Re: Query cancel re-INVITE (Sreenath Kulkarni)
>    5. B2BUA, atomic requests and UPDATE question (Dima Polsky)
>    6. Re: B2BUA, atomic requests and UPDATE question
>       ([EMAIL PROTECTED])
>    7. Re: B2BUA, atomic requests and UPDATE question (Dima Polsky)
>    8. Re: B2BUA, atomic requests and UPDATE question (Paul Kyzivat)
>
>
>----------------------------------------------------------------------
>
>Message: 1
>Date: Tue, 21 Nov 2006 14:24:55 -0000
> From: "Attila Sipos" <[EMAIL PROTECTED]>
>Subject: Re: [Sip-implementors] Maximum size of Header filed Value
>To: "krishna kalluri" <[EMAIL PROTECTED]>
>Cc: [email protected]
>Message-ID:
>       <[EMAIL PROTECTED]>
>Content-Type: text/plain;      charset="iso-8859-1"
>
>
> >>
> >> In my implementation I can't use the dynamic buffers. So I
> >> am trying to think what is the reasonable length of each
> >> header filed value. The total sip message can be 64K that is
> >> fine with me.
>
>Don't consider a "reasonable length of each header field value".
>Instead, just have a scratchpad into which you store strings.
>
>For me, in practice, a 6k scratchpad is large enough to store
>all the SIP strings I need.
>
>Regards,
>
>Attila
>
>
>
> >> -----Original Message-----
> >> From: Attila Sipos
> >> Sent: 21 November 2006 14:20
> >> To: 'krishna kalluri'
> >> Cc: [email protected]
> >> Subject: RE: [Sip-implementors] Maximum size of Header filed Value
> >>
> >>
> >> >>Per line 78 characters + CRLF is the maximum line limit.
> >> By using comma sperated list or line folding you can go up
> >> to 998 characters + CRLF.
> >>
> >> No, you can have more than 78 because it's just a SHOULD,
> >> not a MUST....
> >> (in the real-world I have seen many lines longer than 78 characters)
> >>
> >>    - Lines of characters in the body MUST be limited to 998
> >> characters,
> >>      and SHOULD be limited to 78 characters, excluding the CRLF.
> >>
> >> >>How about if the headers are written like below.  In this
> >> case also the total header field value of Route header is
> >> 998 or Route appears 3 times here and there is no line
> >> folding. So each is 998 characters?
> >>
> >> (I'm not sure I understand you.)
> >> Each line is as long as the text you see
> >>
> >> Regards,
> >>
> >> Attila
> >>
> >>
> >>
> >> -----Original Message-----
> >> From: krishna kalluri [mailto:[EMAIL PROTECTED]
> >> Sent: 21 November 2006 14:09
> >> To: Attila Sipos
> >> Cc: [email protected]
> >> Subject: Re: [Sip-implementors] Maximum size of Header filed Value
> >>
> >>
> >> Hi,
> >> Per line 78 characters + CRLF is the maximum line limit. By
> >> using comma sperated list or line folding you can go up to
> >> 998 characters + CRLF.
> >>
> >> How about if the headers are written like below.  In this
> >> case also the total header field value of Route header is
> >> 998 or Route appears 3 times here and there is no line
> >> folding. So each is 998 characters?
> >>
> >> Route: <sip:[EMAIL PROTECTED]>
> >> Subject: Lunch
> >> Route: <sip:[EMAIL PROTECTED]>
> >> Route: < sip:[EMAIL PROTECTED]>
> >>
> >> In my implementation I can't use the dynamic buffers. So I
> >> am trying to think what is the reasonable length of each
> >> header filed value. The total sip message can be 64K that is
> >> fine with me.
> >>
> >> Thanks
> >> /Krishna
> >>
> >>
> >> On 11/21/06, Attila Sipos <[EMAIL PROTECTED]> wrote:
> >>
> >>
> >> OK, thinking about this again, it looks like that
> >> header field can be any length but....
> >> you can't have any more than 998 characters on one line.
> >>
> >> This does not actually restrict the length of the header because
> >> you can do line-folding.
> >>
> >> So, an example of line-folding:
> >>       Route: <sip:[EMAIL PROTECTED]>, <sip:[EMAIL PROTECTED]>,
> >>              <sip:[EMAIL PROTECTED]>
> >>
> >> But it looks like the actual length in between CRLFs in the body
> >> is restricted to 998.
> >>
> >> Regards,
> >>
> >> Attila
> >>
> >> -----Original Message-----
> >> From: krishna kalluri [mailto:[EMAIL PROTECTED]
> >> Sent: 21 November 2006 10:59
> >> To: Attila Sipos; [email protected]
> >> Subject: Re: [Sip-implementors] Maximum size of Header filed Value
> >>
> >>
> >> Hi,
> >> My question is about the sip header filed value so it maps
> >> to the  "field body" in RFC 2822.
> >>
> >> Eg: Route:<sip:[EMAIL PROTECTED] > ....
> >>
> >> So the value of the Route header filed (any sip header
> >> filed) can be any length or maximum 998 characters.
> >>
> >> Thanks
> >> /Krishna
> >>
> >>
> >> On 11/21/06, Attila Sipos <[EMAIL PROTECTED]> wrote:
> >>
> >> There is no limit in header field size.
> >>
> >> The only limit is the total size of a SIP message which
> >> can be up to 65535 bytes.
> >>
> >> In RFC2822, the 998 character limit is defined in the grammar:
> >>
> >> message         =       (fields / obs-fields)
> >>                         [CRLF body]
> >> body            =       *(*998text CRLF) *998text
> >>
> >> SIP however has no such restriction in the grammar.
> >>
> >> Note also that RFC3261 says:
> >> >> (SIP allows header fields that
> >> >>   would not be valid RFC 2822 header fields, for example.)
> >>
> >> Regards,
> >>
> >> Attila
> >>
> >>
> >> >> -----Original Message-----
> >> >> From: [EMAIL PROTECTED]
> >> >> [mailto:[EMAIL PROTECTED]
> >> Behalf Of krishna
> >> >> kalluri
> >> >> Sent: 20 November 2006 21:08
> >> >> To: [email protected]
> >> >> Subject: [Sip-implementors] Maximum size of Header filed Value
> >> >>
> >> >>
> >> >> Hi,
> >> >> I am reading the RFC 3261 and I have a question about
> >> >> maximum size of header
> >> >> field value.
> >> >> Sip follows RFC 2822. Does this mean the maximum length of
> >> >> the header field
> >> >> value is 998 Characters?
> >> >>
> >> >> Sip header filed values can be written in different format.
> >> >> Here Route header and value was written in 3 ways.
> >> >> Irrespective of the way
> >> >> we write the maximum length of the header filed value is 998
> >> >> characters?
> >> >>
> >> >>       Route: <sip:[EMAIL PROTECTED]>
> >> >>       Subject: Lunch
> >> >>       Route: <sip:[EMAIL PROTECTED]>
> >> >>       Route: <sip:[EMAIL PROTECTED]>
> >> >>
> >> >>       Route: < sip:[EMAIL PROTECTED]>, < sip:[EMAIL PROTECTED]>
> >> >>       Route: <sip:[EMAIL PROTECTED] >
> >> >>       Subject: Lunch
> >> >>
> >> >>       Subject: Lunch
> >> >>       Route: <sip:[EMAIL PROTECTED]>, < sip:[EMAIL PROTECTED]>,
> >> >>              <sip:[EMAIL PROTECTED]>
> >> >>
> >> >> Thanks
> >> >> /Krishna
> >> >> _______________________________________________
> >> >> Sip-implementors mailing list
> >> >> [email protected]
> >> >> https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
> >> >>
> >>
>
>
>
>------------------------------
>
>Message: 2
>Date: Wed, 22 Nov 2006 13:13:30 +0800
> From: "Tang Xi" <[EMAIL PROTECTED]>
>Subject: [Sip-implementors] Query cancel re-INVITE
>To: sip-implementors <[email protected]>
>Message-ID:
>       <[EMAIL PROTECTED]>
>Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>Hi sip group:
>
>I have the following question:
>
>1) UAC send a re-INVITE to UAS
>2) UAS answer a 18x response.
>
>then can UAC send CANCEL to cancel the re-INVITE?
>
>Thanks in advance!
>
>
>------------------------------
>
>Message: 3
>Date: Wed, 22 Nov 2006 00:28:11 -0500
> From: [EMAIL PROTECTED]
>Subject: Re: [Sip-implementors] Re-Invite Cseq
>To: [email protected]
>Message-ID: <[EMAIL PROTECTED]>
>
>    From: Paul Kyzivat <[EMAIL PROTECTED]>
>
>    Brett isn't saying this strongly enough. From 12.2.2 of 3261:
>
>        If the remote sequence number was not empty, and
>        the sequence number of the request is greater than the remote
>        sequence number, the request is in order.  It is possible for the
>        CSeq sequence number to be higher than the remote sequence number by
>        more than one.  This is not an error condition, and a UAS SHOULD be
>        prepared to receive and process requests with CSeq values more than
>        one higher than the previous received request.
>
>And RFC 3261 isn't saying it strongly enough, either:  A UAS *must* be
>prepared to receive and process requests with CSeq values more than
>one higher than the previous received request, if it expects to
>successfully interoperate with UACs.  You can argue that your UAS is
>standard-conformant because the above statement is only a SHOULD
>requirement, but you won't be able to make your UAS work in practice.
>
>Dale
>
>
>------------------------------
>
>Message: 4
>Date: Wed, 22 Nov 2006 11:14:36 +0530 (IST)
> From: Sreenath Kulkarni <[EMAIL PROTECTED]>
>Subject: Re: [Sip-implementors] Query cancel re-INVITE
>To: Tang Xi <[EMAIL PROTECTED]>,       sip-implementors
>       <[email protected]>
>Message-ID: <[EMAIL PROTECTED]>
>Content-Type: text/plain; charset=ascii
>
>Hi,
>
>This is quite possible. You can send Cancel  any INVITE  once u receive  1XX  
>and before receving 2XX response.
>
>/Sreenath
>
>----- Original Message ----
> From: Tang Xi <[EMAIL PROTECTED]>
>To: sip-implementors <[email protected]>
>Sent: Wednesday, 22 November, 2006 10:43:30 AM
>Subject: [Sip-implementors] Query cancel re-INVITE
>
>Hi sip group:
>
>I have the following question:
>
>1) UAC send a re-INVITE to UAS
>2) UAS answer a 18x response.
>
>then can UAC send CANCEL to cancel the re-INVITE?
>
>Thanks in advance!
>_______________________________________________
>Sip-implementors mailing list
>[email protected]
>https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>
>
>
>
>
>
>
>
>__________________________________________________________
>Yahoo! India Answers: Share what you know. Learn something new
>http://in.answers.yahoo.com/
>
>------------------------------
>
>Message: 5
>Date: Wed, 22 Nov 2006 03:14:32 -0600
> From: "Dima Polsky" <[EMAIL PROTECTED]>
>Subject: [Sip-implementors] B2BUA, atomic requests and UPDATE question
>To: <[email protected]>
>Message-ID:
>       <[EMAIL PROTECTED]>
>Content-Type: text/plain;      charset="us-ascii"
>
>Hi,
>
>RFC 3261 states that -  "Note that request processing is atomic.  If a
>request is accepted,   all state changes associated with it MUST be
>performed.  If it is   rejected, all state changes MUST NOT be
>performed. "
>
>Now a B2BUA is described in the RFC as a concatenation of a UAC and a
>UAS. If a call is established through the B2BUA, and an OPTIONS request
>is received by one side of the B2BUA, can it answer the request with a
>final response or should it wait for a response of the other side to the
>forwarded request ? If it waits, can it process additional overlapping
>requests? For example another OPTIONS or INFO?
>
>
>
>Best Regards
>
>Polsky Dima
>
>
>
>
>
>------------------------------
>
>Message: 6
>Date: Wed, 22 Nov 2006 08:17:14 -0500
> From: [EMAIL PROTECTED]
>Subject: Re: [Sip-implementors] B2BUA, atomic requests and UPDATE
>       question
>To: [email protected]
>Message-ID: <[EMAIL PROTECTED]>
>
>    From: "Dima Polsky" <[EMAIL PROTECTED]>
>
>    Now a B2BUA is described in the RFC as a concatenation of a UAC and
>    a UAS. If a call is established through the B2BUA, and an OPTIONS
>    request is received by one side of the B2BUA, can it answer the
>    request with a final response or should it wait for a response of
>    the other side to the forwarded request ? If it waits, can it
>    process additional overlapping requests? For example another
>    OPTIONS or INFO?
>
>It can do any of these, as the SIP transactions on one side of the
>B2BUA are independent (from the SIP standard point of view) from the
>SIP transactions on the other side.
>
>Dale
>
>
>------------------------------
>
>Message: 7
>Date: Wed, 22 Nov 2006 08:10:16 -0600
> From: "Dima Polsky" <[EMAIL PROTECTED]>
>Subject: Re: [Sip-implementors] B2BUA, atomic requests and UPDATE
>       question
>To: <[email protected]>
>Message-ID:
>       <[EMAIL PROTECTED]>
>Content-Type: text/plain;      charset="us-ascii"
>
>Although transactions are independent the info that passes in those
>transactions is related to both UA ( SDP for example, INFO, or query for
>OPTIONS), let's look at mid-dialog UPDATE for a more extreme example -
>it carries a new sdp offer which is of little relevance to the B2BUA but
>should be forwarded from side to side. If the UAS responds to the UPDATE
>request before the UAC received an answer from the other user it may
>create an inconsistent state (if user declines the offer), so should he
>wait?
>
>-----Original Message-----
> From: [EMAIL PROTECTED]
>[mailto:[EMAIL PROTECTED] On Behalf Of
>[EMAIL PROTECTED]
>Sent: Wednesday, November 22, 2006 3:17 PM
>To: [email protected]
>Subject: Re: [Sip-implementors] B2BUA, atomic requests and UPDATE
>question
>
>    From: "Dima Polsky" <[EMAIL PROTECTED]>
>
>    Now a B2BUA is described in the RFC as a concatenation of a UAC and
>    a UAS. If a call is established through the B2BUA, and an OPTIONS
>    request is received by one side of the B2BUA, can it answer the
>    request with a final response or should it wait for a response of
>    the other side to the forwarded request ? If it waits, can it
>    process additional overlapping requests? For example another
>    OPTIONS or INFO?
>
>It can do any of these, as the SIP transactions on one side of the
>B2BUA are independent (from the SIP standard point of view) from the
>SIP transactions on the other side.
>
>Dale
>_______________________________________________
>Sip-implementors mailing list
>[email protected]
>https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>
>
>
>------------------------------
>
>Message: 8
>Date: Wed, 22 Nov 2006 09:39:20 -0500
> From: Paul Kyzivat <[EMAIL PROTECTED]>
>Subject: Re: [Sip-implementors] B2BUA, atomic requests and UPDATE
>       question
>To: Dima Polsky <[EMAIL PROTECTED]>
>Cc: [email protected]
>Message-ID: <[EMAIL PROTECTED]>
>Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>
>
>Dima Polsky wrote:
> > Although transactions are independent the info that passes in those
> > transactions is related to both UA ( SDP for example, INFO, or query for
> > OPTIONS), let's look at mid-dialog UPDATE for a more extreme example -
> > it carries a new sdp offer which is of little relevance to the B2BUA but
> > should be forwarded from side to side.
>
>While what you say may be true for some B2BUAs, it is not something that
>can be asserted generally for all B2BUAs. Apparently you have in mind a
>sort of B2BUA where this is the case.
>
> > If the UAS responds to the UPDATE
> > request before the UAC received an answer from the other user it may
> > create an inconsistent state (if user declines the offer), so should he
> > wait?
>
>3261 doesn't provide any requirements for B2BUAs other than that the
>each side shall follow the rules for a UA. If you want some predictable
>behavior over and above that, then it is up to you, or somebody, to
>specify it.
>
>It sounds like you are talking about the fabled "transparent" B2BUA. But
>AFAIK there are no formal specifications for such a thing. (I guess
>perhaps you could consider some of the IMS specifications to be such,
>but they are very specialized.)
>
>       Paul
>
> > -----Original Message-----
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED] On Behalf Of
> > [EMAIL PROTECTED]
> > Sent: Wednesday, November 22, 2006 3:17 PM
> > To: [email protected]
> > Subject: Re: [Sip-implementors] B2BUA, atomic requests and UPDATE
> > question
> >
> >    From: "Dima Polsky" <[EMAIL PROTECTED]>
> >
> >    Now a B2BUA is described in the RFC as a concatenation of a UAC and
> >    a UAS. If a call is established through the B2BUA, and an OPTIONS
> >    request is received by one side of the B2BUA, can it answer the
> >    request with a final response or should it wait for a response of
> >    the other side to the forwarded request ? If it waits, can it
> >    process additional overlapping requests? For example another
> >    OPTIONS or INFO?
> >
> > It can do any of these, as the SIP transactions on one side of the
> > B2BUA are independent (from the SIP standard point of view) from the
> > SIP transactions on the other side.
> >
> > Dale
> > _______________________________________________
> > Sip-implementors mailing list
> > [email protected]
> > https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
> >
> > _______________________________________________
> > Sip-implementors mailing list
> > [email protected]
> > https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
> >
>
>
>------------------------------
>
>_______________________________________________
>Sip-implementors mailing list
>[email protected]
>https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>
>
>End of Sip-implementors Digest, Vol 44, Issue 39
>************************************************
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to