Re: [secdir] Secdir review of draft-paxson-tcpm-rfc2988bis-02

[Catherine Meadows <[e-mail]>]

I, on the other hand, would say that if the security considerations section is 11 years old,
it is in need of an update or at least another look.  Judging from the remarks that have been
made so far, there is a lot of wisdom about the security impact of RTO that is not captured in the current security considerations section.
At the very least, I think a reference to the security considerations in RFC5681 would be in order.

Cathy

On Apr 21, 2011, at 10:09 PM, Mark Allman wrote:

> 
> I agree with everything Vern said.  But, two more points.
> 
>> Thus, it is a natural target for exploitation for a denial of service
>> attack, in which an attacker convinces a sender to lower its RTO to an
>> unsafe value, causing it to retransmit its packets that are not really
>> lost, and thus lead to congestion.
> 
> First, I don't think this makes sense.  Even if some attacker can
> convince a sender to reduce its RTO and hence trip the RTO early this
> will *reduce* the sender's rate (RFC5681).  That is certainly an
> impairment attack on the connection itself, but that does not "lead to
> congestion".  I.e., it is not somehow an attack on the broader network.
> In fact, the connection would "lead to congestion" with a higher
> probability if it were to continue unimpaired at a higher sending rate.
> 
> Second, I am loathe to change a security considerations section that has
> been good enough for 11 years unless there are actually new security
> considerations.
> 
> allman
> 
> 
> 

Catherine Meadows
Naval Research Laboratory
Code 5543
4555 Overlook Ave., S.W.
Washington DC, 20375
phone: [phone]
fax: [phone]
email: [e-mail]