Message-ID: <23757.1398011078@critter.freebsd.dk>
Date: Sun, 20 Apr 2014 16:24:38 +0000
From: "Poul-Henning Kamp" <phk@....freebsd.dk>
To: discussions@...sword-hashing.net, Bill Cox <waywardgeek@...il.com>
Subject: Re: [PHC] Best use of ROM in password hashing

In message <CAOLP8p7fRf_Z_dC2ccXjzrqu6D8mwPM3j9QPMEWM7PFOQNXzPQ@...l.gmail.com>
, Bill Cox writes:

>Using large ROMs for authentication servers is brilliant.  It is pretty
>hard to steal a TiB or more without being noticed, and if an attacker does
>not have a copy of the ROM, he's stopped cold.

We can argue up and down about how large a ROM you can or cannot
steal, but given that 1TB USB sticks have been available for over
a year, I do not buy the claim that stealing 1TB is "pretty hard".

We still don't know how many GB Snowden got away with and the
*NSA* of all organizations didn't notice a thing...

And a large ROM it's not without its downsides.

You cannot generate that TiB algoritmically and keep the algorithm
around, that would defeat the purpose, it has to be based on
randomness.

That means that you have to back it up, and run integrity tests on it.

Unless you want to dedicate a TB on each and every server, you have
to either make the ROM available across the network or use a centralized
authentication server.

Making it available across the network opens it for theft from any
compromised system, whereas a central authentication server comes
with its own problems in the form of passwords being sent across
networks etc.

Obviously there is a market opportunity here:  Sell batches of data
media with identical copies of a random 1TB ROM.

The only three technologies I can think of which are suitable right
now are USB flash devices, harddisks and optical disks (ie: Sony's
new 1TB "AD - Archival Disk")

All three are at least two orders of magnitude too expensive to be
relevant, all three are subject to wear-out mechanisms, and two
of them probably are prohibitively slow.

That makes a 1TB ROM economically unfeasible, except in installations
which should have known better than to rely on passwords to begin with.

So, yeah, nice in theory, in practice not so much...

That said, an algorithm which can use a ROM sensibly gets a plus
in my book, and another plus if it leaves the size/benefit determination
to the administrator.

Thinking a bit creatively:

A USB stick with an ARM chip can be made for less than $10 in volume.

Putting the entire hashing algoritm on that ARM, including a modest
size random ROM, even as little as 8K, and offloading all password
hashing (but not salt generation) to that ARM chip via the USB port,
means that the ROM can only be stolen by physical access.

And best of all: If you need to hash more passwords than the ARM
will do, you can just plug in another.

If you're worried about loosing track of the USB sticks, you can
give them serial numbers and do a secure transaction to a central
server at startup, to receive the XOR pattern to apply to on-board
ROM before use.

That's probably what I would do, if I were in charge of a big site
relying on passwords today...

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@...eBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.

Powered by blists - more mailing lists