Re: [Idr] IETF LC for IDR-ish document <draft-ietf-grow-bgp-reject-05.txt> (Default EBGP Route Propagation Behavior Without Policies) to Proposed Standard

[Jay Borkenhagen <[e-mail]>]

Eric C Rosen writes:
 > [...]  It would also be nice if the 
 > document indicated that there may be unintended side-effects of changing 
 > the defaults, but the authors consider those to be of no consequence.
 > 

I am not one of the authors (although for what it's worth I have been
agreeing with their comments in this thread), but I don't think
they're saying those side-effects are of no consequence.

While such side-effects may be unfortunate, they will be felt only by
those who do not read release notes, do not test, do not slow-start
their code upgrades, etc.  That such folks may exist should not
prevent standardizing the sane and important defaults this draft
proposes.


To those who propose keeping the ebgp session down if no policies are
configured, that's worse.  It would require an implementation to tear
down an established ebgp session when a necessary policy is removed,
violating POLA.  The actions proposed in this draft constitute failing
hard (hard enough). 

Thanks.

						Jay B.