Re: [Idr] draft-ietf-idr-ix-bgp-route-server-11

[Nick Hilliard <[e-mail]>]

Michael H Lambert wrote:
> This may have been discussed before and I just couldn't find it in
> the archives, and it may be too late in the process to ask the
> question, but should RFC 7606 (Revised Error Handling for BGP) apply
> to peering sessions with route server clients?  It does run counter
> to attribute transparency, but it can help insure that the "right"
> session is reset and that known (or knowable) bad announcements are
> properly propagated.

My reading is that the draft-ietf-idr-ix-bgp-route-server is ok on this.
 The relevant sections in RFC7606 are 7.2 to 7.4, inclusive.  7.3 and
7.4 only make line-protocol statements on what should be considered
malformed.

Section 7.2 states:

>    [RFC4271] also says that an implementation optionally "MAY check
>    whether the leftmost ... AS in the AS_PATH attribute is equal to the
>    autonomous system number of the peer that sent the message". 

I.e. it makes no statement one way or another on what should be done
where the leftmost as != peeras.  draft-ietf-idr-ix-bgp-route-server
doesn't contradict this.

Nick