[extropy-chat] Password Security

[Eugen Leitl]

On Fri, Sep 23, 2005 at 12:36:15PM +0100, BillK wrote:

> <http://www.mandylionlabs.com/PRCCalc/BruteForceCalc.htm>
> "Think your passwords are strong enough to survive a brute force attack?

Some of them, yes. Brute-force attacks on ssh logins are useless (and
there are firewall recipes which actually block attempts by origin after
a few trials -- and of course you can only allow specific IPs in, so
you'd have to know what they are, and to be able to spoof them),
so you actually need access to password shadow file (which is salted), and
this assumes you're already running code on the machine. It is
far easier to just use privilege elevation. The security hole
is somewhere else, and you could drive a truck through it.

> Think again. The keyspace (number of possible combinations) created by
> even the most creative human mind is no match for password audit
> tools.

This is why there methods to automatically generate strong passwords.

pativist
nsessimb
otoricui
atifyinf
alevidop
undepher
ranatent
mardited
skinicar
raldight

might be too easy, as they're pronounceable, and low-caps 
(but remember: they're salted in the system's shadow). How about

shagogastivelay
raninvingeneten
oittidednesslyg
ialentocurugant
pialdevedubragg
blyterialgortia
etinglumbedlyph
thrandurpleress
xtrononvellycle
weebormingiblai

? (Notice: I did not check for properly seeded entropy).

> According to @stake, the Rolls Royce of password auditing tools, their
> LC5 "password auditing tool" includes pre-computed password tables
> containing trillions of password hashes that have been computed in
> advance of the password auditing and recovery process.
> Trillions. That's right, Trillions.

Fiddlesticks. Of what use is your password hash if you don't have
access to the password file? Of which use are hashes, if they're strongly
salted, anyway, and the password is randomly picked from a 26^8 space
(about 10^11, not counting the salt). 
 
> A "strong", humanly generated 8 character password consisting of a few
> upper and lower case letters, a couple of numbers and a special
> character or two approaches approximately only 100 billion
> combinations. Simply put, running a password auditing tool to decode a
> humanly generated password's hash is as fast and automated an exercise
> as spell checking an email."

So don't use humanly generated passwords.

> But while brute-force attacks are easy nowadays, they are rarely
> necessary. Humans are notoriously bad at password security. If you are
> told to create a password, but you must remember it and never write it

Why never writing it down? You'd have to rely on your memory for that.
There is no remote exploit allowing you to read a piece of dead tree
somewhere in my apartment.

> down, what happens? You choose an easy-to-remember password like your
> wife's name or your dog's name. So it is also easy for someone to
> guess. Then you find that you have to remember about twenty passwords
> and never write them down. So you use the same password all the time.
> It's hopeless from a security POV.

Right. So, don't.

In fact, don't use passwords at all. Use smartcards, or crypto tokens.
With PINs. Preferrably, on-token keyboard. That way you never reveal 
your secrets. 
 
> Also people give their passwords to other people. "Just check my email
> for me". People write them down, and other people read them. People
> send them in e-mail, and that e-mail is intercepted. People use them

This is a largely theoretical threat. You have to already 0wn the machine,
or infrastructure downstream. In some cases you have to do a MITM,
as traffic is encrypted, which requires active traffic manipulation
instead of just passive sniffing, a whole different order of magnitude
of attack complexity.

> to log into remote servers, and their communications are eavesdropped
> on. People use public terminals in airports or web cafes and leave all
> their info in the cache when they logoff. Apart from all the
> key-loggers and trojans that are installed on all these public pcs
> (and their own pcs!).

Again, use crypto tokens, with PINs.
 
> The latest trick is man-in-the-middle attacks. Fake bank sites or fake

You misspelled pharming. MITM doesn't mean what you think it means.

> Ebay sites that look identical to the real thing and users happily key
> in all their details.
> 
> Your secrets are NOT safe. Just hope that you never attract the
> attention of a hacker group.

Perfect safety doesn't exist in this universe, but one can come
close, very close. I'm not the right person to do this, but I know
people who can. 

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07100, 11.36820            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20050923/3033b2b9/attachment.bin>