Ben I agree with you about where the weakness lies.  Let me say its
forums like this where we gain the most benefit from those that have
vast experience and we are able to flesh out these issues.  It's a
privilege to be a part of this group.

I understand the encryption part of things.  Encryption is just part of
the whole picture, you don't paint a picture without the paint, and
that's all it is. One color of the picture.  I was just making a case
for the use of a ssl website vs rdp in terms of public exposure.  Same
exposure, same encryption, same problems, but we still do it for the
business need.

Remember who we are talking about, SMB space, until a SMB solution that
does not drive a 10 person or under network over 10k for these "add-ons"
it's simply not going to happen.  They want bang for the buck, and they
surely won't sit down in a sales pitch and let you scare them to death
as to the why you have to do it.  Some will, but most surely do not.

The end point side is absolutely valid and where I personally feel where
80% of most problems are going to be in the near future, if not already.
Problem is that this has the same effect whether you are dealing with
this, a managed VPN, PPTP, IPSEC, whatever the transport is.
Then typically you have HUGE holes open between networks instead of
bringing it down to just one attack surface.

Like you pointed out Ben there is no real 100%, but depending on the
policies in place, (near non existent in SMB's, nor will they pay for it
initially), budget available (hahah, SMB's hardly ever spent budget
dollars just because they may lose it the next year like Corps do) we
work to provide them a relatively secure, cost effective, and value adds
to their business.  As we earn trust and understanding these concepts
are easier for them to digest since we already have a relationship in
place, but it ALWAYS comes down to dollars, never a budget, what is
leaving my wallet as a result of what you are saying, and how is it
going to put money back in to replace it or prevent against more
leaving.  Arguably the last point is where I stand my ground, but
sometimes it's not enough.

They are just different waters we navigate in the same big ocean.  

Greg

-----Original Message-----
From: Ben Scott [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, April 01, 2008 5:34 PM
To: NT System Admin Issues
Subject: Re: Public TS - opinions?

On Tue, Apr 1, 2008 at 5:09 PM,  <[EMAIL PROTECTED]> wrote:
> Correct me if I am wrong, but isn't RDP 128 bit encrypted ...

  Gah.  Encryption isn't like horsepower.  Comparing key sizes doesn't
tell you how secure something is.  And you can't make something secure
just by sprinkling a little cryptology on it.

  In particular, the size of the session key doesn't matter when
there's no host authentication at all, and user authentication is done
via a password that's probably got about 10 to 20 bits of entropy if
you're lucky.  And is probably the user's last name if you're not.
Google "man in the middle attack" for just one scenario.

  But even then, I'd be a lot more worried about the end-points.  Most
compromises have come from the end-points, not the transport.  We're
talking about allowing connections from unknown, untrusted, arbitrary
clients.  Most of which probably already have some kind of malware on
them.  Keystroke loggers would be especially scary.  And can the
secondary channels that RDP uses for things like remote printing act
as a conduit for malware to the server?

> 6.       Token, 3rd party authentication

  Aren't two-factor authentication products also rather expensive?
Certainly more-so than, say, OpenVPN, or even the IPsec stuff built-in
to Windows, no?

  That said, the rest of your recommendations are certainly good; I
just think that's not where the weaknesses lie.

-- Ben

~ Upgrade to Next Generation Antispam/Antivirus with Ninja!    ~
~ <http://www.sunbelt-software.com/SunbeltMessagingNinja.cfm>  ~

~ Upgrade to Next Generation Antispam/Antivirus with Ninja!    ~
~ <http://www.sunbelt-software.com/SunbeltMessagingNinja.cfm>  ~

Reply via email to