Re: [NSE script] SSH weak hostkey

[Brandon Enright <bmenrigh () ucsd edu>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, 08 Sep 2008 18:59:04 -0300
Arturo 'Buanzo' Busleiman <buanzo () buanzo com ar> wrote:

> Fyodor wrote:
> > Hi Sven.  Sounds like a great script, but unfortunately 8MB is too
> > large to ship with Nmap by default.  Though people could download
> > and
>
> I've just finished this first test, anyone care to implement it into
> the NSE script?
>
> http://tools.buanzo.org/sshvulnkey/$type/$key
>
> where $type = 'dsa1024' | 'rsa2048'
> and $key = last 20 chars of a key, like '0011223344aabbccddee'
>
> The php script returns '0' or '1'.
>
> Anyone care to test it?

Hi Arturo,

I think the trouble with a service like this is that it by necessity
reveals weak keys semi-publicly.  This makes me nervous for the DNS
check against DNS-OARC (and I trust them) but it makes me _really_
nervous for something like weak SSH keys.

Suppose you see some box at UCSD check your service for thousands of
keys and it turns out that hundreds of them are weak?  The logical
conclusion would be that UCSD is a great place to SSH scan for weak
keys.

It also makes is possible to compare competitors.  Suppose 30% of the
keys a Bank of America scanner checks are weak but Citi Bank scanner
only has 1%.  Maybe Citibank isn't such a great place to do business.
The fact that this data *might* be recorded by some third party makes
companies really nervous.

It really only takes one tech writer to freak out and right a headline
like "Nmap leaks hacker information to Argentina company".  I think you
are trustworthy but rolling this service up into an NSE script that
ships with Nmap forces users to make the same trust assessment.

Nevermind that *you* aren't going to use the data (or probably even
record it).  I think that enough paranoid people will freak out about
this sort of service that we need to tread very lightly in this area.

I guess that was more than my fair share of $.02 .  Please don't take
the above as a personal attack but I think it is something worth
considering.

Brandon

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)

iEYEARECAAYFAkjFqssACgkQqaGPzAsl94L6uACgtAaosCO/wT7bxDQyne2LsCnP
I5UAn1bqDUVIIENRcKVQsgXwjV5Wl45w
=wkPs
-----END PGP SIGNATURE-----

_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org