[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[AVT] RE: [PWE3] Re: I-D ACTION:draft-ash-avt-ecrtp-over-mpls-protocol-00.txt
Stewart and all,
I am not sure I understand the idea about the new registry.
ECMP safety of the "generic" CW is based on the fact that
value 0 is not used in the "protocol versions" registry, and
the decision to swap the CW and RTP header in TDM PWs over MPLS
has been based on value 8 (that is what the nibble in question is
in the RTP header) has been allocated for something called PIP in this
registry even if it probably has never been used.
If we go for the new registry, what is going to happen if somebody,
at some time, decides to allocate value 4 (or, in fact, any value between 4
and 9) in it? Do we expect IANA to maintain the linkage between the
two seemingly unrelated registries?
With best regards,
Sasha Vainshtein
email: sasha@axerra.com <mailto:sasha@axerra.com>
tel: +972-3-7659993 (office)
+972-8-9254948 (res.)
+972-58-674833 (cell.)
> -----Original Message-----
> From: Stewart Bryant [mailto:stbryant@cisco.com]
> Sent: Monday, February 16, 2004 2:00 PM
> To: Curtis Villamizar
> Cc: avt@ietf.org; pwe3@ietf.org
> Subject: [PWE3] Re: I-D
> ACTION:draft-ash-avt-ecrtp-over-mpls-protocol-00.txt
>
>
>
> There are some operations that need to be able to distinguish
> between IP and non-IP MPLS payloads. For example a router
> performing ECMP based on the IP addresses or other information
> in the IP header, or an IPFIX/PSAMP packet sampling system
> that needs to analyse the payload contents.
>
> We identified this need in PWE3 and proposed a solution. This
> is described in sections 5.4.3 and 5.4.4 of the PWE3 Architecture
> draft:
>
> http://www.ietf.org/internet-drafts/draft-ietf-pwe3-arch-06.txt
>
> In essence we proposed that the nibble that immediately follows
> the MPLS bottom label has the following semantics:
>
> Value Meaning
> 0 PWE3 payload*
> 1 Identified payload type (see below)
> 4 IPv4
> 6 IPv6
>
> * It might be argued that this should be changed to mean
> "unidentified payload type"
>
> The identified payload type above is the first nibble
> in a longword with the following definition:
>
> 0 1 2 3
> 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6
> 7 8 9 0 1
>
> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
> |0 0 0 1| reserved = 0 | PA | Protocol ID
> |
>
> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
> | As defined by PPP DLL protocol definition
> |
> |
> |
>
> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>
> Figure 14: PWE3 PID
>
> The meaning of the fields of the PWE3 PID (Figure 14) is
> as follows:
>
>
> PA protocol authority for the user plane or the
> control plane
> protocol ID
> 0 = PPP DLL
> 1-15 = Reserved
>
> Protocol ID
> Protocol ID following the format defined by the protocol
> authority identified in PA.
>
> Bits 4 to 11 inclusive are reserved for future use and
> must be zero.
>
> PWE3 currently describes this in terms of historic IP version numbers,
> but IESG feedback is that we need to use a new registry.
>
> The proposal is that there is a new "nibble that follows the MPLS
> label" registry. This will still have the values and
> semantics described
> above. A new draft describing this will be submitted when the IETF
> embargo on new drafts lifts, and the PWE3 architecture draft modified
> accordingly. Note that this does not change the existing PWE3 protocol
> or headers, just the way that we describe the protocol.
>
> If this proposal gains approval, the implication for ecrtp-over-mpls
> would be that it should either:
>
> 1) Run under first nibble 0
> 2) Run under first nibble 1 with an ecrtp-over-mpls as the
> PPP-DLL value
> 3) Request a first nibble value from the new registry
>
> - Stewart
>
>
> Curtis Villamizar wrote:
>
> > ------- Blind-Carbon-Copy
> >
> > To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
> > cc: avt@ietf.org
> > Reply-To: curtis@fictitious.org
> > Subject: Re: I-D
> ACTION:draft-ash-avt-ecrtp-over-mpls-protocol-00.txt
> > In-reply-to: Your message of "Mon, 09 Feb 2004 07:39:42 CST."
> >
> <9473683187ADC049A855ED2DA739ABCA0201F057@KCCLUST06EVS1.ugd.att.com>
> > Date: Mon, 09 Feb 2004 11:28:58 -0500
> > From: Curtis Villamizar <curtis@workhorse.fictitious.org>
> >
> >
> > In message
> <9473683187ADC049A855ED2DA739ABCA0201F057@KCCLUST06EVS1.ugd.att.com>
> > , "Ash, Gerald R (Jerry), ALABS" writes:
> >
> >>All,
> >>
> >>Please review and comment on 'Protocol Extensions for ECRTP
> over MPLS' http:/
> >>/www.ietf.org/internet-drafts/draft-ash-avt-ecrtp-over-mpls-
> protocol-00.txt.
> >> The protocol extensions enable the use of MPLS to route
> enhanced compressed
> >>RTP (ECRTP) compressed packets over an MPLS LSP without
> compression/decompres
> >>sion cycles at each router.
> >>
> >>The proposed extensions are based on the requirements
> documented in 'Requirem
> >>ents for ECRTP over MPLS'
> http://www.ietf.org/internet-drafts/draft-ash-avt-e
> >>crtp-over-mpls-reqs-01.txt.
> >>
> >>Thanks,
> >>Jerry
> >>
> >>
> >> Title : Protocol Extensions for ECRTP over MPLS
> >> Author(s) : J. Ash
> >> Filename : draft-ash-avt-ecrtp-over-mpls-protocol-00.txt
> >> Pages : 0
> >> Date : 2004-2-5
> >
> >
> >
> > Jerry,
> >
> > This looks to be a very good direction to go in. I've dropped MPLS
> > from the Cc and put MPLS on the bcc so people know the discussion is
> > on the AVT mailing list only.
> >
> > In your encapsulations you should make sure to avoid any
> values in the
> > first byte that could be mistaken as IPv4 or IPv6 or any of the
> > encapsulations being worked on by those working on VPN and PW. See
> > "The PWE3 Control Word", Jonathan Stein, 22-Oct-02,
> > draft-stein-pwe3-controlword-00.txt, for some early discussion on
> > that. Note though that "Payload Type" is something that many
> > providers require so that multipath can be used on IP traffic which
> > consitutes the majority of traffic yet reordering is avoided for
> > traffic which cannot be split based on what appears to be
> an IP header
> > and the IP src/dst pair.
> >
> > It might be best to reserve a single fixed first byte for
> the "Packet
> > Type" (and I suggest you make up a name for this shim such as
> > SCID_Packet_Type) and use the second byte for the packet types that
> > you have defined. Just as you have to get numbers for SCID_Request
> > Object and Header_Compression_Reply Object from IANA, you'll have to
> > coordinate with others over this number space, though it is
> less clear
> > if IANA is the authority on this quite yet.
> >
> > Curtis
> >
> > ------- End of Blind-Carbon-Copy
> >
> >
> >
>
>
> _______________________________________________
> pwe3 mailing list
> pwe3@ietf.org
> https://www1.ietf.org/mailman/listinfo/pwe3
>
_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt