Re: [secdir] Secdir last call review of draft-ietf-rtgwg-spf-uloop-pb-statement-09

[Stewart Bryant <[e-mail]>]

On 07/01/2019 16:11, Phillip Hallam-Baker wrote:
> Reviewer: Phillip Hallam-Baker
> Review result: Has Issues
>
> The document describes the problem and solution pretty clearly. Unfortunately,
> there is no discussion of the security considerations which is not appropriate
> for a document addressing an availability which is a security issue.
>
> While microloops can form by chance, some consideration should be given to the
> possibility that an attacker could induce a loop to perform a DoS attack.

In section 1 the text says:

[RFC8405] defines a solution that satisfies this problem statement
    and this document captures the reasoning of the provided solution.

It is safe to assume that the reader of this text would have read 
normative reference RFC8405 and thus would be fully aware of the 
security issues related to the solution being analysed.

An attacker that had access to a network such that they could induce 
microloops would have the ability to do many worse things to the network.

If they were able to attack in-band they could poison the routing system 
to take it down in far more interesting ways. Operators use security at 
the physical and network layer to prevent this.

If they were operating at the physical layer then they could take 
circuits down at will and cause microloops in the base protocol, traffic 
overloads and application malfunction.

Thus if the attacker could deploy either of those attacks in a network 
to induce micro-loops, then any security considerations in this draft 
would count for nothing.

The draft is an analysis, and thus I think that it correctly states that 
it introduces no additional matters for security consideration.

- Stewart