On Monday, April 30, 2012 2:51:20 PM UTC+5:30, Dag-Erling Smørgrav wrote:
> Harry <
simon...@gmail.com> writes:
> > Given,
> > 1. that Alice is about to ssh to Bob;
>
> let's say "Bob's machine", since SSH is not a person-to-person protocol.
> > 2. that Mallory possesses Alice's private key; and
> > 3. that Alice is on a Linux machine;
> >
> > how can Mallory sniff the traffic between Alice and Bob?
>
> He can mount a man-in-the-middle attack and hope that Alice doesn't
> check the fingerprint before accepting the server's host key and that,
> if Bob's machine's host key is already listed in her known_hosts file,
> she not only ignores her SSH client's warning about duplicate keys but
> also removes the existing entry before trying again.
>
> In other words, exactly what he would do if he didn't have Alice's
> private key. Alice's private key is only used for authentication, not
> for encryption, so he doesn't gain much from knowing it - except for the
> ability to log in on Bob's machine as Alice, which may or may not enable
> him to snoop on Alice, depending on whether (and how well) Bob's machine
> is hardened and on what Alice uses SSH for.
>
> > If necessary, we can also assume,
> > 4. that Mallory also has root access to Alice's machine.
>
> That's easy. He can replace the SSH client with one that logs
> everything, or just snoop Alice's tty.
>
> DES
> --
> Dag-Erling Smørgrav -
d...@des.no
Using opensource tools, I'd like to carry out an actual (ethical) MITM attack between two hosts that I control. Where can I get more, hands-on type of info about this?
Although the subject of this post mentions 'ssh', I'd like the equivalent info for SSL as well.
I don't mind writing some brief code in C/Perl if necessary, but I'm hoping I will not have to understand all the gory details of SSH and SSL protocols.