mailing list archives
Re: Risk measurements
From: Valdis.Kletnieks () vt edu
Date: Fri, 12 Feb 2010 21:32:32 -0500
On Fri, 12 Feb 2010 16:54:48 +0100, Christian Sciberras said:
And who do you know what the bugs are? Risk modeling cannot solve this
kind of issue. Vulnerabilities aren't intentional.
It isn't intentional that I could piggyback a particular process and
get kernel access. Since vulnerabilities are based on exceptions, how
do you know that this kind of exception occurs?
Again, mathematics lose ground here.
Actually, it turns out that you can do most of this without having a *clue*
what the bugs are. It's counter-intuitive, but true.
Let's say we have a server that handles SSN data, and has a trust relationship
with another server that handles Visa cards. Neither of those servers have
any connection to the server in R&D that has product design data.
Now we can say with a high degree of certainty that if that server gets
whacked, we have a high probability of SSN exposure, potential exposure of Visa
card numbers, and essentially zero R&D exposure. Dig a bit further - the server
runs Apache, OpenSSH, and PHP. OpenSSH is firewalled to only a section of the
corporate network. We have a pretty good handle on how often Apache, OpenSSH
and PHP get whacked (advisories per year is a pretty good place to start). We
can now model things like "how likely an Apache hole will end up with us
leaking Visa cards", "how likely an OpenSSH 0-day will pwn our R&D", and so on.
I don't need to be able to predict the next bug. I only need to be able to
predict "*BSD and OpenSSH have a good track record, so they'll probably keep
being reasonably trustable". I also don't need to be able to guess *what*
the next bug in phpnuke will be - I feel pretty safe in predicting that if
the bozos in Advertising insist on installing it on an outward-facing server,
we'll have an incident within the year.
-"Unfortunately, you'll need to do some risk modeling to figure out
what "reasonable bounds" is for each piece of information."
Wait, so I need to do risk modeling to quantify the risks of
information/results of a risk assesment on software? Sounds like
beauroucracy to me (pun intended).
No, you made it too complicated. Lose the second "of a risk assessment".
"You need to do risk modelling to quantify the risks of information/sofware."
I see the reason behind risk management, but I don't see it being
usefull except in policy-making.
That's because most of its value *is* in policy-making and related decisions
for implementing the policy. "We're in good shape on this system, Payroll
needs some defense-in-depth, and we need to either buy bullets for those bozos
in Advertising or KY". That's risk management in one sentence. ;)
Full-Disclosure - We believe in it.
Hosted and sponsored by Secunia - http://secunia.com/
Re: SMS Banking Craig S. Wright (Feb 10)