Re: [Int-area] 答复: 答复: 答复: Is the UDP destination port number resource running out?// re: I-D Action: draft-ietf-intarea-gue-04.txt

[Tom Herbert <[e-mail]>]

On Sat, May 20, 2017 at 4:16 PM, Joe Touch <[e-mail]> wrote:
>
>
> On 5/19/2017 11:09 PM, Xuxiaohu wrote:
>
> GUE is supposed to be both signalling and content (data), where the data are
> IP
> packets.
>
> Since IANA strives to assign one port for a service, IP packet within the
> UDP tunnel should be assigned a dedicated port. In other words, GUE and
> IP-in-UDP are distinguished by the different port numbers.
>
> GUE is one service that includes both encapsulation of IP packets and
> signaling.
> Frankly, it seems like it would work anywhere IP works - where IPv0 is
> defined as the signalling channel (which is sufficient because IPv0 isn't
> defined).
>
> In that case, the first field after v0 needs to be a signal channel version
> number, to allow for future updates.
>
>
> Take away the IP part and GUE isn't an E anymore.
>
> Services are expected to have version fields and subtype
> demultiplexing indicators, to so that all message variants of current
> and future versions can use a single port number.
>
> Sure, the version field within the IPvx packet could be used for
> demultiplexing
>
> purpose.
>
> That demultiplexes within IPvx. There still needs to be a way to demultiplex
> non-IPvx packets (control) from IPvx.
>
> Since GUE and IP-in-UDP have different UDP port numbers,
>
> They don't and they shouldn't. That would complicate forwarding - a single
> service needs to use a single port. Using separate ports complicates
> configurations - this is a case where you want "fate sharing" (either both
> IP encapsulation and the signal channel work or neither do).
>
> I don't know why there is still a need to demultiplex GUE and IP-in-UDP.
>
> The point of GUE is an IP encapsulation channel with in-band signalling.
> That is a single service, IMO.
>
> Note - AFAICT, GUE could work anywhere an IP packet works. IP packets always
> start with a version number, and v0 isn't really defined. Defining v0 as the
> signal channel is the same thing as how GUE is currently specified.
>
Joe, I'm not sure I understand your point about IPv0. GUE v0
encapsulates packets of any IP prolocol number (IP, GRE, MPLS,
EtherIP. ...) and allows for data messages as well as control
messages. GUE v1 directly encapsulates IPv4 and IPv6 in a form of
header compression. Where would we use IPv0?

Tom

> Joe
>