Terminology for Benchmarking BGP Device Convergence in the Control Plane
RFC 4098

Note: This ballot was opened for revision 06 and is now closed.

(Alex Zinin) Yes

(Harald Alvestrand) No Objection

Comment (2004-04-01 for -)
No email
send info
Reviewed by Lucy Lynch, Gen-ART

This document does not use 2119 language. There are places where the word "should" is used, but in the context, one would think that the meaning of 2119 "MUST" is the intended reading.
Still, the doc seems good. Some more nits in the review (spelling sized).

(Steven Bellovin) No Objection

Comment (2004-10-13)
No email
send info

(Margaret Cullen) No Objection

(Ted Hardie) No Objection

Comment (2004-03-29 for -)
No email
send info
Section 2.2 has this definition:

      The network prefix length is the number of bits out of the total
      available in the address field, used to define the network prefix.

I hate to nit-pick definitions in a terminology documents, but I
had to read this over and over before I could make this mean
what I think they mean.  I take it they mean the network prefix
length describes the number of bits used by the network prefix
out of the total available in the address field; the placement
of "total available" in the above makes that hard to infer.


"The network prefix length contains the number of bits used
by the network prefix out of the total available in the address

make sense as a replacement?

NIT:  DUT is first expanded as "Cevice Under Test"

NIT:  Thus they will forward every packet to a specific next hop
      based on the longest match between  the  destination IP addresse
      and the routes in the forwarding table.

-->IP address

A quiet rant on section 4.1.3:

(No change needed, but this does work my nerves enough that it ought
to be fed back to the authors)

      A full default free table is the union of all sets of default free
      BGP routes collectively announced by the complete set of
      autonomous systems making up the public Internet.  Due to the
      dynamic nature of the Internet, the exact size and composition of
      this table may vary slightly depending where and when it is

While I am sure that the authors understand far better than I do
how path-vector protocols work, I am concerned that the use of
this phrase without comment may reinforce a common misunderstanding.
Folks new to this (and even some folks with considerable operational experience)
often use phrases like this in ways which elide the critical fact that each route implies
a particular place in the topology.  The set created by this
union is very different both in size and character from the set created
by the union of all reachable destinations, because each route is
a unique "how do I get there from *here*" instead of a unique "there".
If the authors could make that point with a hammer, this phrase might
not go the way of "global routing table" as a term of art.

(Scott Hollenbeck) No Objection

(Russ Housley) No Objection

(David Kessens) (was Discuss, Yes) No Objection

(Allison Mankin) No Objection

(Bert Wijnen) No Objection