Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 13 Dec 2012 21:22:50 -0800
From: Steven Alexander <>
Subject: Re: Intentionally Increasing Collisions in Password
 Hashing Algorithms


So that's a longer way of saying that I really hope I'm not wasting
> everyone's time.

Quit worrying about that.  One of the benefits of entertaining crazy ideas
is that it can help us to look at things from new angles.  The
discussions/analysis could end up being valuable even if your idea isn't.

Then there's the basic question about how much is it worth to an
> attacker to be able to break into an account?

That really depends.  For a banking/payment website, the account itself has
value based on what the attacker can extract from it.  For other sites, the
value of the accounts probably comes from being able to log in with the
same information elsewhere.  The attacker doesn't care about your actual
Gawker account but if he can use it to log in to your bank, PayPal or WoW
account, it has value.  Of course, this hits on what you were trying to
accomplish with the truncated hashes.  If the Gawker account can't be used
to predict one of those others, it has no value.

I don't know if an attacker would be interested in having a small number of
low value accounts.  It probably depends on how much effort is required to
turn one into a high value account.  If 10% of Gawker accounts can be
converted into a higher value account with just a few tries at different
sites, that gives them some potential value.  If the success rate is .01%,
then the individual value is low by any measurement.

I'd be interested to know more about how stolen accounts are actually used
and traded.  How often are the initial attackers actually using the
accounts and how often are they selling them to a third party?  Are the
attackers and sellers both involved in trying to compromise additional
accounts with known credentials or is it primarily one or the other?

"How many guesses can an attacker make?"

A lot of sites have lockout policies.  If you lock the account after 3
tries or 10 tries, the access-by-collision issues becomes almost moot (on a
per account basis anyway).  Even with temporary lockouts or delays that
limit the attacker to 100-300 guesses per day, having a chance of collision
between 1/2k and 1/8k could be acceptable.

The better question might be, "how many guesses *do* attackers make?"  The
GW2 article said that they were saying targeted guessing with only one to a
few guesses per account.  If that's the case, then the idea of truncating
isn't helpful; you need blacklisting and/or two factor authentication.

BTW, I read your "Testing Metrics..." paper.  Have you done any testing to
see how well blacklisting would help against offline attacks?


Content of type "text/html" skipped

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ