Re: [Idr] comments on draft-ietf-idr-rs-bfd

[Gert Doering <[e-mail]>]

Hi,

On Wed, Aug 02, 2017 at 11:07:45AM +0200, Robert Raszuk wrote:
> Interesting ...
> 
> Look at section 5 of yr draft:
> 
> "The next hop field is the key for the NH-Reach NLRI type".
> 
> As defined today your new NLRI encoding in MP_REACH does not have per
> client distinguisher. And if it is the NLRI key as the draft states good
> luck to have the next hop information it carries scoped to be per client.

It's an implemention decision on whether the RS sees this as global 
or per-client thing.

Since RSs already today do many things in ways the original BGP specs
did not envision (like, per-client bestpath calculation) doing this 
particular one globally does not sound (as said before) like the most
reasonable implementation decision.

OTOH, if the draft had a few more words on this (like, "It is RECOMMENDED
that next-hop reachability is considered a per-client value.  A route-server
MAY decide that a given next-hop should not be considered anymore for
any client if more than a configurable threshold of clients report 
unreachability.") it would help to de-confuse readers.

Gert Doering
        -- NetMaster
-- 
have you enabled IPv6 on something today...?

SpaceNet AG                        Vorstand: Sebastian v. Bomhard
Joseph-Dollinger-Bogen 14          Aufsichtsratsvors.: A. Grundner-Culemann
D-80807 Muenchen                   HRB: 136055 (AG Muenchen)
Tel: [phone]           USt-IdNr.: DE813185279