Re: [Bier] Comments on draft-chen-bier-frr-02.

[Tony Przygienda <[e-mail]>]

On Wed, Mar 17, 2021 at 8:14 PM Huaimo Chen <[e-mail]>
wrote:

> Hi Tony,
>
>     Thanks much for your comments.
>     My responses are inline below with [HC].
>
> Best Regards,
> Huaimo
> ------------------------------
> *From:* Tony Przygienda <[e-mail]>
> *Sent:* Wednesday, March 17, 2021 12:47 PM
> *To:* Jeffrey (Zhaohui) Zhang <zzhang=[e-mail]>
> *Cc:* Huaimo Chen <[e-mail]>; [e-mail] <
> [e-mail]>
> *Subject:* Re: [Bier] Comments on draft-chen-bier-frr-02.
>
>
>
>
>
> In my view, the following does handle multiple failures to maximum extent:
>
>
> zzh> A better way is to simply have a single BIFT for both normal and FRR
> forwarding. Each entry has some ECMP and/or primary/backup branches,
> and you simply use another ECMP branch or a backup branch when the
> normally used branch fails. A backup branch can go through completely
> different nbr, or go through the same neighbor but via a tunnel.
>
> Consider that and the cost associated with building/maintaining multiple
> BIFTs, I would argue that the multiple per-nbr FRR BIFT approach, while
> perhaps acceptable for a particular implementation, is not ready yet for WG
> adoption.
>
>
>
> Please note that I am not arguing against LFA based FRR. I am just saying
> that the reference model of a single BIFT with ECMP/primary/backup branches
> for an entry is the best baseline.
>
>
> as participant and from experience
>
> +10
>
> in that sense merling's paper seems a better model to talk about stuff.
> Plus the technical concerns about the compressed multiple BIFTs on multiple
> failures
>
> [HC]: The starting point of the draft is the LFA-based BIER FRR.
> The compression is optional. Regarding to multiple BIFTs or a single BIFT,
> we can include a single BIFT in the draft for comparisons.
> Since the draft is experimental, we can choose one after
> some experiments and/or research on them.
>
>
Yes, I think it's a good idea. Having both models in the draft will make
this a framework for BIER FRR [without TE] (including original work by
Menthe on single BIFT with LFA failovers). Addressing Jeffrey's technical
comments with experiments/further evolution is a good idea as well. And
adding clear sections on the use cases/feasiblity and deployment model in
case of IGP, no IGP signalling.

As you say, experiments in implementation should give us some kind of
comparison and real world input on the feasibility, +/- of different
approaches

thanks

-- tony

-- tony