Re: [spring] Penultimate Segment Popping and RFC8200 (Was Re: We don't seem to be following our processes (Re: Network Programming - Penultimate Segment Popping))

[Robert Raszuk <[e-mail]>]

Hi Brian,

I think everyone grasps it. I think the problem is that the SRH proponents
> use the word "insert" a little ambiguously. In IPv6-land, it's been assumed
> to mean "insert an SRH header in the middle of an existing IPv6 packet
> header".
>

Well I got that. In fact I recommend to some friends to rename it in the NP
draft to "impose" or "place" so the text does not raise blood pressure
where it does not to :)

But again as you see from the subject line this discussion is about
deletion not imposition nor insertion. I bet 6man list will be the forum to
discuss Daniel's draft soon.


> I suspect that in SRH-land, and in the IPv6 data plane context, it's mainly
> being used to mean "encapsulate an IPv6 packet in a new IPv6 packet that
> includes
> an SRH header."


I think in the context of SRv6 it was/is used to mean both. Insert SRH with
new header or without one. And the context which exact action is required
is inferred from the situation.

Yes this adds unnecessary confusion and I am sure editors will be happy to
clear this in documents.

Said all this NP draft under this discussion does not have any issue with
any of this.


> And in the MPLS data plane context it means "prepend an additional
> MPLS label."
>

> Right or wrong?
>

Well in MPLS label stack insertion was more called label imposition.

Thx a lot,
R.