On Wed, 25 Nov 2015, Paul Jakma wrote:
If it goes past MED it may well just hit #10, 'prefer older', which depends
on received order again. (That's actually pretty evil and probably also a
malfeature).
Oh, I'm all in favour of a consistent decision process, especially one
that defines a total order over paths based on information in the route.
DMED doesn't give that though. Always-compare would though, and makes
sense as a post-AS_PATH/steer-equal-length metric.
I'm starting to think "we" in the BGP engineering world need to consider
just deprecating MED from BGP altogether, and define it as exclusively for
the always-compare, intra-AS/steer-equal-as-path-len use-case.
regards,
--
Paul Jakma p...@jakma.org @pjakma Key ID: 64A2FF6A
Fortune:
I stopped believing in Santa Claus when I was six. Mother took me to
see him in a department store and he asked for my autograph.
-- Shirley Temple
_______________________________________________
Quagga-dev mailing list
Quagga-dev@lists.quagga.net
https://lists.quagga.net/mailman/listinfo/quagga-dev
