Re: [sidr] AD Review of sidr-origin-validation-signaling-09

["John G. Scudder" <[e-mail]>]

On Nov 30, 2016, at 9:18 AM, Randy Bush <[e-mail]> wrote:
> section 4.5 of 4593 is relevant, or all of sec 4

Thanks, used in the text below.

> i am kinda sad that 7132 is not too good on this

I looked there first but it's a *path* security threat model so can't really be blamed for not covering this.

Candidate new security section below. I'd appreciate an ack from Alvaro that this addresses his concern before I publish.

--John

6.  Security Considerations

   Security considerations such as those described in [RFC4272] continue
   to apply.  Since this document introduces an extended community that
   will generally be used to affect route selection, the analysis in
   Section 4.5 ("Falsification") of [RFC4593] is relevant.  These issues
   are neither new, nor unique to the origin validation extended
   community.

   The security considerations provided in [RFC6811] apply equally to
   this application of origin validation.  In addition, this document
   describes a scheme where router A outsources validation to some
   router B.  If this scheme is used, the participating routers should
   have the appropriate trust relationship -- B should trust A either
   because they are under the same administrative control or for some
   other reason (for example, consider
   [I-D.ietf-sidr-route-server-rpki-light]).  The security properties of
   the propagation path between the two routers should also be
   considered.  See [RFC7454] Section 5.1 for advice regarding
   protection of the propagation path.

(all the refs above are in the "informative" section)