Last Call Review of draft-ietf-mpls-in-udp-04
review-ietf-mpls-in-udp-04-opsdir-lc-brownlee-2014-01-09-00

Request Review of draft-ietf-mpls-in-udp
Requested rev. no specific revision (document currently at 11)
Type Last Call Review
Team Ops Directorate (opsdir)
Deadline 2014-01-16
Requested 2014-01-02
Other Reviews Genart Last Call review of -04 by Roni Even (diff)
Genart Last Call review of -09 by Roni Even (diff)
Secdir Last Call review of -09 by Charlie Kaufman (diff)
Secdir Early review of -03 by Charlie Kaufman (diff)
Review State Completed
Reviewer Nevil Brownlee
Review review-ietf-mpls-in-udp-04-opsdir-lc-brownlee-2014-01-09
Posted at http://www.ietf.org/mail-archive/web/ops-dir/current/msg00100.html
Reviewed rev. 04 (document currently at 11)
Review result Has Nits
Draft last updated 2014-01-09
Review completed: 2014-01-09

Review
review-ietf-mpls-in-udp-04-opsdir-lc-brownlee-2014-01-09

Hi all:

I have performed an Operations Directorate review of
  draft-ietf-mpls-in-udp-04

  "This document specifies an IP-based encapsulation for MPLS, called
   MPLS-in-UDP (User Datagram Protocol)."

  It also explore the use cases for such a protocol.

- - - -

1. Is the specification complete?  Can multiple interoperable
     implementations be built based on the specification?

Yes.

2. Is the proposed specification deployable?  If not, how could it be
     improved?

Yes.

3. Does the proposed approach have any scaling issues that could
     affect usability for large scale operation?

I guess that the most likely use of this protocol will be by ISPs
who need to make an MPLS tunnel via the Internet to a small remote
site - that is, it's unliekly to be deployed on a large scale.

4. Are there any backward compatibility issues?

No, this is a new, specialised, tunnelling protocol.

5. Do you anticipate any manageability issues with the specification?

Since it's most likely to be deployed by ISPs, they'll need to
make sure that their PE equipment implements it correctly, so that
it interworks as expected with the MPLS systems at each tunnel endpoint.

As for manageability, it introduces an unreliable (UDP) link into
an MPLS network - that means ISP operators will need to be aware
of such links so that they can minimise the time taken to diagnose
link failure, or sudden increases in packet loss rate.

6. Does the specification introduce new potential security risks or
     avenues for fraud?

Yes; the drafts Security Considerations cover this well.

Two suggested improvements:

- Section 3, the packet layour diagram shows the UDP Destination
  port name as 'MPLS.'  Using 'MPLS-in-UDP' would save people like
  me betting this confused with the port for passing MPLS LDP info.

- Section 3: Destination port number - maybe add 'or MPLS-in-UDP
  with DTLS' as you do in section 4.


Cheers, Nevil
Co-chair, IPFIX and EMAN WGs

--
---------------------------------------------------------------------
 Nevil Brownlee                          Computer Science Department
 Phone: +64 9 373 7599 x88941             The University of Auckland
 FAX: +64 9 373 7453   Private Bag 92019, Auckland 1142, New Zealand