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

[Job Snijders <[e-mail]>]

On Fri, Apr 21, 2017 at 11:35:06AM +0200, Robert Raszuk wrote:
> > Did you review Alvaro's suggested changes which prompted the third IDR
> > consultation on this topic? They are viewable here:
> >
> >     http://htmlpreview.github.io/?https://github.com/jaredmauch/
> > draft-mauch-bgp-reject/blob/alvaro/draft-ietf-grow-bgp-
> > reject-06-from-5.diff.html
>
> Do we really have consensus among all authors of this document (leave
> alone rest of IETF) that it should apply to all AFI/SAFIs ?

Feel free to double check with the GROW chairs or the assigned OPS Area
Director.

> Last ... what's your take on multiple comments regarding build in
> silency of this ?
> 
> See session will be established just fine ... even incoming prefix
> count will be incremented just fine.

The perceived 'silency' is a misnomer. Vendors are free implement
feedback mechanisms (such as a syslog or terminal warning: "you did not
configure a policy"), and some already provide feedback like this.
Nothing in this draft precludes vendors from ensuring the operation of
their BGP implementations a pleasant experience.

No feedback is required for the other side of the EBGP session, after
all the ownership of the change event lays with the side that upgrades
their software (same applies when the 'changing side' removes a policy,
or applies a deny-all policy, or any other standard operating
procedure).

You also fail to acknowledge that today's deployed BGP speaking software
presents a myriad of behaviours: some already follow bgp-reject, some
apply the concept in one direction and some operate in an promiscuous
insecure mode. This draft provides guidance to unify those behaviours
and provide a safe, consistent experience across multiple
implementations. 

> Last time I checked until you display bgp table there is no separate
> counter in any implementation to indicate in BGP summary how many out
> of received paths were best path eligible.

I encourage you to continue your studies on how today's routers work and
what counters are available.

Kind regards,

Job