[secdir] Secdir review of draft-ietf-roll-trickle-mcast-05

[Tero Kivinen <[e-mail]>]

I have reviewed this document as part of the security directorate's
ongoing effort to review all IETF documents being processed by the
IESG.  These comments were written primarily for the benefit of the
security area directors.  Document editors and WG chairs should treat
these comments just like any other last call comments.

This document describes the Multicast protocol for Low and Lossy
Networks. This protocol uses trickle algorithm. I am not familiar
enough to trickle to really analyze what the protocol does. Security
considerations section mentions that the protocol uses sequence
numbers to keep track of messages, and attacker who can insert
messages can mess up with those sequence numbers, and attacker can
then flush messages from the buffered messages list, and can also
allow setting it high enough so recipients will not get any messages
as they have too small sequence number.

The protocol has no protection against this attack, but notes that
both of those are denial-of-service attacks and devices can protect
against them by using link-layer security mechanisms. It also claims
that those mechanisms are typically employed without specifying which
security methods it is pointing to. I do not know how often those
link-layer security methods are really used. Perhaps it would be
useful to list some of those security methods here.

I do not have any other comments for the protocol, and otherwise I
think the document is ready, but as I said I did not have time to
really analyze the protocol itself.
-- 
[e-mail]