All,

Looking at all the drafts mentioned I do think there are actually two
different scenarios being considered and perhaps it is actually beneficial
to treat them separately.

Original draft allowed to signal link bw in situations where you peer from
different boxes. Path hiding is not an issue. Most likely such signalling
is to be done in respect to underlay.

However newer proposals aim to enhance the overlay behaviour where via one
or the other technique path hiding issue of BGP is addressed.

Inter-AS option C is also an overlay so I would recommend to think twice
before influencing base idea with this deployment scenario.

To conclude let's also realize that link bw is actually not a property of a
path as it is a property of the next hop binded to such path (how easy or
difficult it is to get to such next hop). And here I must say that the
draft Kaliraj & Minto proposed draft-kaliraj-idr-multinexthop-attribute
could possibly be used as a vehicle for its encoding.

The clear benefit for such perspective is ability to signal helper next
hops addresses with their link bw metrics possible hashing spread without
need to explode BGP with lot's of paths respectively if that is done by
add-paths or RD prepend.

Thx,
Robert


On Mon, Jun 26, 2023 at 11:44 PM Alvaro Retana <[email protected]>
wrote:

> On June 26, 2023 at 5:27:39 PM, Jeff Tantsura wrote:
>
> Jeff:
>
> > Over the last couple of years I have reached out to the authors of the
> > original draft at least twice with a request to refresh the draft and
> bring
> > the necessary changes in, without much success though.
> >
> ...
> > Note that there’s also an EVPN specific draft (standard track).
>
> That made me go look -- I found two related documents!  Are these what
> you're referring to?
>
>    https://datatracker.ietf.org/doc/html/draft-ietf-bess-weighted-hrw
>    https://datatracker.ietf.org/doc/html/draft-ietf-bess-evpn-unequal-lb
>
>
> > I’d be happy to volunteer to help managing the mess ;-)
>
> I know the idr/bess-chairs are listening. ;-)
>
>
> Alvaro.
>
> _______________________________________________
> Idr mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/idr
>
_______________________________________________
BESS mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/bess

Reply via email to