Re: [Bier] WGLC: draft-ietf-bier-php

["Jeffrey (Zhaohui) Zhang" <[e-mail]>]

Hi,

> In my understanding, from the control plane point of view, the LSP does not end on the penultimate node, the destination is the ultimate node. While in the forwarding the bier header could be popped on the penultimate node and packet sent to the ultimate node, but the LSP's destination still is the ultimate node.

That's how I see it as well. In BIER case, the penultimate node is a BFR, and the ultimate node is a BFER in control plane (in that it has BFR-id and participate in BIER signaling) though its penultimate node will pop the BIER header in the data plane.

Jeffrey


Juniper Business Use Only

-----Original Message-----
From: BIER <[e-mail]> On Behalf Of Dhruv Dhody
Sent: Wednesday, August 19, 2020 10:24 AM
To: Loa Andersson <[e-mail]>
Cc: Gengxuesong (Geng Xuesong) <[e-mail]>; [e-mail]; Greg Shepherd <[e-mail]>
Subject: Re: [Bier] WGLC: draft-ietf-bier-php

[External Email. Be cautious of content]


Hi Loa,

On Wed, Aug 19, 2020 at 5:07 PM Loa Andersson <[e-mail]> wrote:
>
> All,
>
> yes, support.
>
> I have one thought that possibly don't need to have an impact on 
> progressing this draft.
>
> If you think of the p2mp forwarding tree in bier as a p2mp LSP, it 
> seems to me that the LSP setup by this draft ends on penultimate node 
> (which is kind of an oxymoron). The ultimate node is also the node th 
> do the PHP, which would make it UHP. If the LSP ends on a node that is 
> the ultimate node.
>
> For example in MPLS the LSP ends on the ultimate node, while the 
> pen-ultimate node does the PHP.
>
> It would also be interesting to know how PCE does this. Dhruv?
>

In my understanding, from the control plane point of view, the LSP does not end on the penultimate node, the destination is the ultimate node. While in the forwarding the bier header could be popped on the penultimate node and packet sent to the ultimate node, but the LSP's destination still is the ultimate node.

BTW in PCE P2MP documents, you would not find any reference to PHP.
>From the P2MP tree path computation point of view, the computation is
done till the ultimate node and it is up to the signaling to handle PHP.

Thanks!
Dhruv

> As I said, don't let thgis stop progressing this draft, but we maybe 
> should sit down and and come with a unified way of using the concepts.
>
> On 19/08/2020 09:40, Gengxuesong (Geng Xuesong) wrote:
> > Yes, support !
> >
> > Best Regards
> >
> > Xuesong
> >
> > *From:*BIER [mailto:[e-mail]] *On Behalf Of *Greg 
> > Shepherd
> > *Sent:* Thursday, August 13, 2020 11:31 AM
> > *To:* [e-mail]
> > *Subject:* [Bier] WGLC: draft-ietf-bier-php
> >
> >
> > Please read and respond to this thread. We had a WGLC for this draft 
> > previously, but had no response. Let’s do better.
> >
> > https://urldefense.com/v3/__https://datatracker.ietf.org/doc/draft-i
> > etf-bier-php/__;!!NEt6yMaO-gk!UBxdbv0B2SFGtXr2xVzNSjDXpbhg6p9WGswOPp
> > SndIsFhVWf0FSRu27sTq60eF-O$
> >
> > Start of short LC timer. Respond by Aug 19..
> >
> > - Shep
> >
> > (Chairs)
> >
> >
> > _______________________________________________
> > BIER mailing list
> > [e-mail]
> > https://urldefense.com/v3/__https://www.ietf.org/mailman/listinfo/bi
> > er__;!!NEt6yMaO-gk!UBxdbv0B2SFGtXr2xVzNSjDXpbhg6p9WGswOPpSndIsFhVWf0
> > FSRu27sTlE8dlHC$
> >
>
> --
>
> Loa Andersson                        email: [e-mail]
> Senior MPLS Expert                          [e-mail]
> Bronze Dragon Consulting             phone: [phone]

_______________________________________________
BIER mailing list
[e-mail]
https://urldefense.com/v3/__https://www.ietf.org/mailman/listinfo/bier__;!!NEt6yMaO-gk!UBxdbv0B2SFGtXr2xVzNSjDXpbhg6p9WGswOPpSndIsFhVWf0FSRu27sTlE8dlHC$