Early Review of draft-ietf-idr-aigp-14
review-ietf-idr-aigp-14-opsdir-early-bonica-2014-01-23-00

Request Review of draft-ietf-idr-aigp
Requested revision No specific revision (document currently at 18)
Type Early Review
Team Ops Directorate (opsdir)
Deadline 2014-04-22
Requested 2014-01-09
Authors Prodosh Mohapatra , Rex Fernando , Eric C. Rosen , Jim Uttaro
I-D last updated 2014-01-23
Completed reviews Genart Last Call review of -16 by Dan Romascanu (diff)
Genart Telechat review of -17 by Dan Romascanu (diff)
Opsdir Early review of -14 by Ron Bonica (diff)
Opsdir Last Call review of -16 by Stefan Winter (diff)
Assignment Reviewer Ron Bonica
State Completed
Request Early review on draft-ietf-idr-aigp by Ops Directorate Assigned
Reviewed revision 14 (document currently at 18)
Result Has issues
Completed 2014-01-23
review-ietf-idr-aigp-14-opsdir-early-bonica-2014-01-23-00
Folks,

I have been asked to provide an OPS-DIR review of draft-ietf-idr-aigp. Reading
the draft, I have one question and one comment:

Question
========
Would it be a good idea for BGP speakers to negotiate  Aggregated IGP Metric
capability at session initiation?

The text in Section 3.3 helps somewhat. But wouldn't it be neater to deal with
configuration mismatches at session initiation time than to ignore mismatches
later in the life of the session?

Comment
=======
The draft might benefit from an Operational Considerations Section. In that
section, you might address a scenario in which an AS1 peers with AS2. AS1
contains a relatively small number (<100) of IGP links, but many (>400,000) BGP
routes. In AS1, one link flaps slowly. This causes the Aggregated IGP Metric to
change for ~100,000 BGP routes and lots of BGP activity between AS1 and AS2.

What steps should be taken at AS1? At AS2?

--------------
Ron Bonica