Slashdot Mirror
Calculating the Cost of Full Disk Encryption
CowboyRobot writes "Is full disk encryption (FDE) worth it? A recent study conducted by the Ponemon Institute shows that the expected benefits of FDE exceed cost by a factor ranging from 4 to 20, based on a reduction in the probability that data will be compromised as the result of the loss or theft of a digital device. 'After doing all of the math, Ponemon found that the cost of FDE on laptop and desktop computers in the U.S. per year was $235, while the cost savings from reduced data breach exposure was $4,650.'"
The TCO of is more than the cost of installing it.
No sig today...
One issue with IT security is that policies and security measures like this are only one small part of the picture. My partner works in a government affiliated company and has to use FDE for all PCs. Because of how they have implemented it they virtually all still use the default key (which wasn't random) and if you change it then you thwart the original intent of having quasi-hotdesks.
Passwords written on desks, stuck to the screen etc are common in many places. Sending files off-network to places like dropbox or email to get around security 'hassles' is widespread. The owner of my current firm wants to use an iPad, because we won't let it on network he does most email from a web email account!
FDE with rubbish passwords is entirely pointless as anyone with motivation to get in can. If you start requiring complex passwords the risk of people writing it down and storing it with the PC increases. We need to stop thinking about security as a technical issue and work out how to produce 'secure enough' systems which users don't subvert or misuse.
If removing security breaches is worth nearly $5k a year then surely using some kind of RFID security card that must be near the PC/Laptop to unlock would be cost effective. I could keep it in my wallet or as a keyring. Even better would be combine it with a RSA style password device for two-factor authentication when providing a password (thus making less complex passwords less of an issue).
In a corporate environment, you have to have some kind of key management system. You can't do FDE with a free utility that is just "Enter the password to get in to the computer." Well why not? Tow big reasons:
1) What if the person suddenly up and dies, and you need to get at the data? A backup won't help if said backup is also encrypted with the same password that only they knew. You need to have a system to get in.
2) More commonly, what do you do when a user forgets their password? This happens ALL the fucking time. People cannot remember passwords, just how it is. Just losing data is not an acceptable answer, so you have to have a system that can get in.
Now there are systems out there like that. They have central key stores, key recovery facilities and so on all while maintaining cryptographic security. However all the ones I've seen cost money. Then on top of that is the cost of administering such a system.
As an example at work a lady forgot her password, as she is known to do on days ending in "y". So she couldn't get in the encrypted laptop that has key codes for the doors (she deals with that). She also hadn't put the laptop on the 'net in like a year, so it was all desync'd with the Active Directory. This meant my boss couldn't log in to do any kind of override. So he had to hook it up, go through this key recovery thing where the console give you a bigass key to enter in to the system, then get it to sync passwords, then he could log in and get everything working. Took a fair bit of time to do.
You have to count all that kind of thing in cost calculations. You can't pretend like it isn't a cost. Yes you already pay his salary but he has about 5,000 other thing to be doing that weren't being done while he worked on that. Needless to say if this were being used for more than a couple systems (we only use it in special cases) it would quickly need one or more people who's job was to administer it and deal with all the problems caused by it (meaning by users).
That's kind of a flawed way to look at things imo (no offence intented).
Having 99 locks on the frontdoor is indeed pointless, but that's because all those locks perform the same function. Using (for example) a virusscanner, a firewall and full disk encryption is more akin to having both a lock on the frontdoor but also a fire extinquisher in the hallway. It's a very sane and generally smart thing to have.
On a sidenote, having 2 (different) locks on a door may very well be a smart thing, because a Type A lock has different flaws than a type B lock. Having 2 different kinds means you potentially eliminate a variety of easy exploits that target a specific type of lock. I don't know where the "line" is but i find it hard to imagine more than 2 or 3 locks being a sane scenario for "regular" entrances.
Now it's obviously true that every security measure brings with it a "penalty" if you will, to legit users. For most locks this means the user has to carry a key on him, something most people in the west consider normal and not a high price to pay. Having to remember a PIN code to use your debit card is also a penalty on legit accountholders, but we accept it. Now wether or not full disk encryption is as pricey as the article seems to make out, i dunno, but lets look at the alternatives.
A) Unencrypted drives
B) Certain Encrypted drives/containers/shares/whatever
C) Full Disk encryption
In this day and age (dare i say it, the Information Age) there is a definite demand for encrypting sensitive and important data for almost everyone but the most untrained pc user. This makes option A a bad choice for almost everyone which means the decision is not just about "should we use FDE or not", it becomes "do we encrypt EVERYTHING or just cetain objects" and in order to make that determination you need to look at the price for both.
Let's not forget, simply enabling FDE is a minor act for the sysadmin whose installating the system, and since FDE is transparent to userspace applications, there is very little additional configuration or problems that arrise due to it. It will impact performance of the hardware, however one can take this into account during the purchase of new machine(s) and simply take a slightly faster system to make up for it (if that's even needed, usually it ain't). Only encrypting certain partitions, shares or folders means you have to deal with many additional administrative and training issues. How do you prevent ppl from moving sensitive files from the secure X:\ drive to the C:\ drive, how do you teach the (presumably non-IT) users where they can store which files. Which policies have to be enforced to ensure it all stays on the level. How often (if at all) do you audit the systems for "leaked" senstive files, etc.
I'm not a sysadmin myself, but i do work in IT and have a fair understanding on their job, and to me at a quick glance it appears FDE is a LOT cheaper and easier for everyone involved than encrypting only sensitive files.