Re: Adoption Call for "The IPv6 Compact Routing Header (CRH)"

[Robert Raszuk <[e-mail]>]

Hello Ron,

I recommend you take a look at this URL:

https://mailarchive.ietf.org/arch/msg/ipv6/Ef05LFFij45mm8fM8hLFXknxoIA/

with two additional observations:

#1 - Reference can be encoded into a new fixed and immutable RH (applied on
end hosts or ingress routers) ways easier than a chain of CRIDs.
Literally from IPv6 side you are not "encoding" the pointer into the
address. You are simply specifing next loopback address which is of 128
bits. Plain IPv6. Packets arriving at that loopback have mapped
instructions how to forward them.


#2 - For cases discussed recently where some folks are planning to use CRH
over Internet I am afraid a lot of third party networks may just blindly
fail to deliver such packets:
https://tools.ietf.org/html/draft-ietf-opsec-ipv6-eh-filtering-06
Therefore for such use case I do recommend use of RbR.

Best regards,
R.



On Wed, May 27, 2020 at 5:32 AM Ron Bonica <rbonica=
[e-mail]> wrote:

> Mach,
>
> The CRH is unique in the following respect. It does not rely on an
> instruction or a path being encoded in the IPv6 Destination Address. It
> relies only on RFC 4291 IP Address semantics.
>
> Can you show me another IPv6 traffic steering solution that:
>
> - is equally compressed
> - does not rely on an instruction being encoded in the IPv6 destination
> address
> - does not rely on all endpoints being numbered from the same IPv6 prefix
>
>
>                         Ron