Tero, Dan and others,

I've read your informational draft.

Overall comments:

  The draft contains quite a lot of background information (what you are
  trying to achieve on technical point of view, what were the
  alternative solutions considered). Part of this background is also
  available on WESP draft.

  Making this draft an information disclosure on "algorithm to
  determine if IPsec ESP packet stream has been encrypted or not",
  without too much explanation or hand waving would increase its
  usability. The background could be find by-reference on the WESP
  RFC.

  Please consider adding definitions/glossary entries for the
  following concepts: flow, flow-cache. I know they are relevant on
  certain implementations, but not necessarily well defined on that
  sense, or at least introducing these terms properly before using
  them.

About the abstract:

  Consider changing abstract in a way that really points out the
  good on this approach. Something like:

-8<---

  This document describes an algorithm for distinguishing IPSEC
  ESP- NULL packets from encrypted ESP packets.  The algorithm can
  be used on intermediate devices, like traffic analyzers, and deep
  inspection engines, to quickly decide whether given packet flow is
  interesting or not. Use of this algorithm does not require any
  changes made on existing RFC4303 compliant IPSEC hosts.

-8<---

About the algorithm:

  Pseudo-code seems sane, and was understandable after reading the
  specification part. However I did not try to implement it, nor
  really proof its correctness.

P.S. Please ignore the following meaningless whining:

About Security Considerations:

  Raises a question, why is this all needed? What is the use case?
  Neither this, nor WESP draft can give a simple answer for the
  question. If encrypted traffic is allowed to pass traffic
  analyzer/intrusion detector/intrusion prevention/firewall/whatever,
  the malicious end nodes will use encryption. If encrypted traffic
  is dropped by intermediate, then there is an end node policy
  issue.  Detection whether an ESP packet payload is encrypted or
  not should already be fast on deep inspection packet matching
  hardware.

--
Tero Mononen <[email protected]>
_______________________________________________
IPsec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipsec

Reply via email to