[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Reasons for AH & ESP

To: sankar@nexsi.com
Subject: Re: Reasons for AH & ESP
From: Paul Koning <pkoning@xedia.com>
Date: Fri, 2 Jun 2000 18:04:54 -0400 (EDT)
Cc: ipsec@lists.tislabs.com
References: <Pine.OSF.4.10.10005281122430.29651-100000@sparrow.qut.edu.au><00060214380700.00924@odin>
Sender: owner-ipsec@lists.tislabs.com

>>>>> "Sankar" == Sankar Ramamoorthi <sankar@nexsi.com> writes:

 Sankar> I am not sure if any one is using AH and ESP this way, but
 Sankar> this is one reasoning I was given some time back.

 Sankar> RFC 2401 section 4.3 talks about combining security
 Sankar> associations (iterated tunneling). It talks of a case (case
 Sankar> 2) where an end-point could apply ESP with the outer gateway
 Sankar> and AH with a host behind the gateway. Thus the packet could
 Sankar> be authenticated and encrypted over the internet and just
 Sankar> authenticated inside the network behind the gateway allowing
 Sankar> for any traffic analysis.

I think you lose confidentiality doing this, because it leaves the
packets vulnerable to Steve Bellovin's splicing attack on ESP.  To
defeat that attack, you have to do authentication in the same box that
does the decryption.

     paul

Follow-Ups:

Re: Reasons for AH & ESP

From: Sankar Ramamoorthi <sankar@nexsi.com>

References:

Re: Reasons for AH & ESP

From: Sankar Ramamoorthi <sankar@nexsi.com>

Prev by Date: Re: Interoperability (was: Death to AH?)
Next by Date: Re: Reasons for AH & ESP
Prev by thread: Re: Reasons for AH & ESP
Next by thread: Re: Reasons for AH & ESP
Index(es):
- Main
- Thread