audit perspective: proof that all connections are encrypted

From: Florin Andrei (florin_at_andrei.myip.org)
Date: 09/16/05

  • Next message: David Wolever: "ssh -R only listening on lo" 
    To: secureshell@securityfocus.com
Date: Thu, 15 Sep 2005 16:05:38 -0700

    I have what's perhaps a slightly unusual question.

    Suppose company X is going through an audit (think: SOX). Suppose one of
    the questions that the auditors ask is: "we want proof that all your
    remote access devices only allow encrypted connections, not plaintext".

    With a VPN concentrator, that's easy: you show them the encryption
    algorithms that are enabled, show them that plaintext is a disabled
    option and they're happy.

    But how about openssh? Which is the config item in sshd_config that says
    "if the client does not agree with all these encryption schemes, all of
    which are not plaintext, terminate the connection"?

    Essentially, we have to prove that plaintext is rejected by the server.

    Any connection with the Ciphers and MACs options in sshd_config?

    Hopefully I'm making myself understood. This is not a strictly technical
    question, it's somewhere on the border between technical issues and
    legal issues. I need an answer that will satisfy people who are not
    geeks - if I'm being sent in the right direction I can build a coherent
    response myself (hopefully) but I need a starting point.

    I believe that this kind of issue will become more common in the near
    future, as the practice of auditing will extend to more and more
    companies.

    Thanks, 

    -- 
Florin Andrei
http://florin.myip.org/
  • Next message: David Wolever: "ssh -R only listening on lo"

    Relevant Pages

    • Re: Backup of encrypted Data in the Cloud
      ... Unix timestamp at encryption time and the secure hash of the plaintext ... that there is none.If it is secure, then using the timestamp is overkill. ... Thus I need a second value apart from the plaintext hash, here the timestamp at encryption time fit in quite well. ...
      (sci.crypt)
    • Re: Modelling a Cipher and Creating its Random Keyset.
      ... What do you call a plaintext? ... ?The range of the change-of-origin array therefore has to be ... encryption algorithm having been validated by that algorithm, ... ?This is a de facto random keyset that is in the ...
      (sci.crypt)
    • Indistinguishability and integrity in symmetric encryption
      ... "The 'right' security property for general-purpose symmetric encryption". ... >symmetric encryption scheme (for which the empty plaintext is not ... A has interfaces to an encryption oracle ... It is assumed that the ciphertext returned by A is different to all those ...
      (sci.crypt)
    • Re: NEWBIE QUESTION: Key space exhaustion - How do I know when Im there?
      ... key exhaustion is the only option, ... looking at the result of a block decryption, then by all means you have ... found a big fat weakness in the encryption system. ... E.g. the resulting plaintext "makes sense" ...
      (sci.crypt)
    • RE: Encrypted Communications and Predictable Communications?
      ... There are more sophisticated attacks which might use known plaintext ... consider a database system, with a client application on one machine and a database server on another. ... How much does the predictability of such message exchanges ... Should the encryption system take steps to ensure that the encrypted data contains random information to pad out messages to at least the minimum ...
      (SecProg)