Re: Evaluation: draft-ietf-idmr-igmp-mrdisc - Multicast Router Discovery to Proposed Standard

[Russ Housley <housley@vigilsec.com>]

Russ Housley        [   ]     [ X ]       [   ]      [   ]

Comment:

I think that the Security Considerations section is pretty weak.  I do not 
feel strongly enough to delay the document, but I want to document my 
concerns.  I have two points.

First, if a rogue sends a Multicast Router Advertisement message, then 
legitimate routers will send multicast traffic to the associated port. The 
authors correctly point out that this is an easy way to 
eavesdrop.  However, further explanation is needed to understand how these 
can become a denial of service attack on legitimate multicast flows.

Second, the authors assert that the messages are extensible, and thus 
security features could be added later.  While I see how this is possible 
for data integrity and authentication, I do not see how this is possible 
for confidentiality.  I do not see the benefits of encryption in this 
protocol, but the authors raise it as a possibility in the future.  What is 
the threat that they are thinking about?

Russ