Skip to main content

Early Review of draft-ietf-idr-flowspec-redirect-ip-10
review-ietf-idr-flowspec-redirect-ip-10-secdir-early-sethi-2026-05-03-00

Request Review of draft-ietf-idr-flowspec-redirect-ip
Requested revision No specific revision (document currently at 16)
Type Early Review
Team Security Area Directorate (secdir)
Deadline 2026-04-16
Requested 2026-04-01
Requested by Susan Hares
Authors Jeff Haas , Wim Henderickx , Adam Simpson
I-D last updated 2026-08-13 (Latest revision 2026-05-21)
Completed reviews Rtgdir Early review of -06 by Ines Robles (diff)
Opsdir Early review of -06 by Linda Dunbar (diff)
Secdir Early review of -10 by Mohit Sethi (diff)
Genart IETF Last Call review of -10 by Elwyn B. Davies (diff)
Comments
Please examine the implementation reports at
https://wiki.ietf.org/group/idr/implementations/draft-ietf-idr-flowspec-redirect-ip
and within another draft using this draft:
https://datatracker.ietf.org/doc/draft-ietf-idr-ts-flowspec-srv6-policy 
(see section 7).
Assignment Reviewer Mohit Sethi
State Completed
Request Early review on draft-ietf-idr-flowspec-redirect-ip by Security Area Directorate Assigned
Posted at https://mailarchive.ietf.org/arch/msg/secdir/PtHw56w_WsuOg6AgtgigoPFb4_w
Reviewed revision 10 (document currently at 16)
Result Ready
Completed 2026-05-03
review-ietf-idr-flowspec-redirect-ip-10-secdir-early-sethi-2026-05-03-00
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 is naturally a bit tough to grasp for someone without routing
background. However, it essentially defines a new BGP Flow-spec action that
allows a BGP peer to request specific traffic be redirected or copied to a
target IP address (presumably a specialized inspection server(s)). The primary
security risk with this mechanism is unauthorized traffic manipulation.
Specifically, a peer attempting to hijack traffic that doesn't belong to them,
or redirecting it to a target IP they don't control. To mitigate this, the
document relies on a two validations. First, it specifies that baseline
Flow-spec validation (RFC 9117) SHOULD be applied to verify the peer has
legitimate authority over the traffic pattern being filtered. Second, it
introduces a new rule stating that receiving peers must validate that the
announcing peer also holds legitimate routing authority over the specific
target IP address. This ensures a peer cannot manipulate traffic unless they
own both the original destination and the redirect target. Perhaps the
requirements language can be carefully evaluated. For example: "[RFC9117]
SHOULD be applied" vs. "SHALL" or "MUST".