Slashdot Mirror
Storage Security
Storage Security is not about turning on the right configuration options on your XYZ brand server appliance. It's about applying solid, methodical security practices to your storage systems, regardless of whether they are disks directly attached to a single computer, Network Attached Storage or part of a Storage Area Network. The authors address the full security cycle, too, starting with evaluating the security of proposed new storage solutions. Comparative data in hand, the book shows you how to narrow the field to a single solution that offers the best balance between functionality and security.
And once the system is selected, you can't stop there. You've got to decide on appropriate security policies for the new storage system, draft and implement a backup and restore plan, deal with disaster recovery and take care of a host of other issues. In short, this is a good guide to an entire range of considerations necessary to select, deploy and manage a secure storage solution.
The book's evaluation methodology is particularly valuable. Each type of storage (directly attached, NAS and SAN) is covered in a chapter of its own. Within each chapter, the authors address specific technologies used to implement that type of storage. For example, the direct-attach chapter discusses such common storage technologies as SCSI and IDE, moderately exotic systems like USB and Firewire drives, and some more advanced solutions like HiPPI and SSA. Each technology is then placed in a matrix and scored in 11 different categories, including popularity and industry acceptance, built-in data protection features, typical fault tolerance and physical security characteristics.
The authors assign each rating on a scale of 1 (poor) to 5 (the best). This gives a good general indication of how each technology measures up, but they tend to rely on a straight average of the ratings when determining the best technology. Although it's true that the average allows you to make a quick ballpark comparison, there are many other factors to consider as well, such as the suitability for your particular environment and the way in which your users need to access their data. The matrixes are quite useful, but just remember that you can't always boil things down to a simple numerical score.
Probably the biggest problem with this book is that it's pretty dry. As a reference book, the writing style is fine, since it's easy to find what you're looking for, and the chapters are concise. It's difficult to read from cover-to-cover, though, which is a shame because that's what you should probably do the first time through. Take it in small doses, a chapter or so at a time, and you should be fine.
Storage Security is about just what you'd think: the security of your data as it's being stored on your server(s). It's not a detailed look at the configuration of any one product, but rather a comprehensive, theory-based approach to managing the security of your storage subsystem from evaluation to purchase to daily operations. If you manage a small or mid-size network, you may or may not need this book. If you have a larger network, though, or have significant data-storage needs, this deserves a space on your shelf.
You can purchase Storage Security: Protecting, SANs, NAS and DAS from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
Duct tape your storage devices.
Protect Freedom!
Buy Tom Ridge Brand Duct Tape, it's minty fresh!
This is something I've wondered about, and this reminded me.
Is there any operating system that supports encrypting the storage at the file system level. In other words, is there any operating system where I can specify that I want a particular path encrypted, and then copy files over, edit files, etc without worrying about having to decrypt, recrypt manually all the time?
Even a weak encryption would satisfy me.
This is nothing new. Administrators and other have known for a long thime that no machine is secure if someone has access to the physical machine. If someone can open your box up, reboot it, attach new devices to its private subnet, etc. then it is not secure.
Why anyone thinks this should be different for storage systems is beyond me. If someone can break in and steal your data, or attach new devices to the data transfer channel, or whatever, then your data isn't secure. That the authors think this comes as a revelation to anyone means either they are really stupid or they think administrators are really stupid.
Get your machines installed in a location with good physical security. That's really all there is to it.
"Security is a process, not a product"
Why is this something you'd need a book for? It comes down to the basics.
One, never allow physical access to what you're trying to secure.
Two, _never_ allow physical access to what you're trying to secure.
Three and so on: log all security events, break users into groups for simplification of manageability, set permissions properly, protect network shares and device access, and be aware of the content that's being secured (what it is, how it's used, etc.)
Beyond that, there's not much else to consider. However, from the review it looks like they go beyond security issues to stuff like, "what hardware is best for my particular application." Sure, it's a big consideration. For example, you wouldn't want a two million row database running off a Network Appliance over NFS across switched fast ethernet, but there's enough free information out there that making decisions like that should be trivial. Furthermore, if you're doing system specifications w/o knowing about the technologies you have to choose from, I sure hope you're not an employee of the place for which you're building a system, because they're not going to like you very much when it dies on a regular basis.
Not having actually read the book, I may be way off base, but from the title and the review above, I don't see how it would be all that helpful except maybe as a primer to a junior-level engineer.
IEEE is working on a standard (IEEE P1619) which will allow encryption to shared media (SANs I guess). They've set up a working group.
They're looking at (from the website):
Something to look at in the future. Hopefully it'll be more secure than WEP. :)
MS products, in particular, like to create a large number of temp files and there is no way of configuring where they are kept. I'm not sure if OSS alternatives have this configuration ability.
And of course, you also have to worry about elements of documents in memory (which can be recovered).
I've had this similar thinking before, because the information in and of itself is not important, from a technical perspective, it's the mechanism to access that needs to be secure. Hence, a SAN with a fibre-channel fabric would seem secure (a client needs an HBA card), but hook it up to a MS File server, SQL Server, or Oracle, and suddenly all the same exploits apply.
;)
I would suggest it's not the type of nails used, it's the design of the front door. I could be wrong.
Peace, Out!
~Airrage
"This isn't a study in computer science, its a study in human behavior"
When we talk about "Information Assurance," it's based on 3 principles:
Confidentiality--Nobody reads your data unless they're allowed to (think top-secret information)
Integrity--Nobody can change your data unless they're allowed to (think bank account balance)
Availability--When you need the data, it's there (backups, redundancy, etc.)
I do what the voices on my console tell me to do.
No one but me seems to use them, but I've personally seen amazing reliability with magneto-optical (MO) drives and media. In fact, I've never had one of my MO disks fail. I certainly can't say that about other media I've used:
- casettes (TRS-80 model 1 circa 1977)
- floppies (actually, 8" floppy disks are very reliable, and they go down from there)
- various kinds of streaming tape
- Bernoulli discs of various capacity
- zip discs of various capacity
- hard drives of various capacity
- CD-R and CD-RW of various capacity
Seriously, consider MO media such as the Fujitsu 1.3 gig discs and drives. Of course this does not address really long-term storage and the issues of lost/failing hardware and standards over decades, but I think this gives you one of the most stable physical formats available for "near-line" storage.
One particular memo was about the servicing of the water coolers, and went out to the whole company. When I strings'ed the memo, though, at the top was a draft of an employee's termination letter! Oops. Apparently, this was the undo buffer for Word -- the writer of the memo had just written the termination letter, printed it, deleted it from the document, and wrote the water cooler memo in the same instance of word. However, if opened this doc in Word, I couldn't access the hidden info, no matter what I tried.
Since then, I've always wondered how much other sensitive information has snuck out, via MS Word.
Keys are stored in smart cards. Reading backup tapes requires a Cyberntics drive and the correct key. Obviously you need to manage this very well to avoid being SOL during an actual recovery situation.
Of course, consider how vulnerable your backup media really is. I don't need to hack your network, just show up in an Iron Mountain uniform with forged ID maybe 30 minutes before the real Iron Mountain guy shows up. I then drive off with ALL you data.
If your children ever found out how lame you are, they'd murder you in your sleep
I used to work for a Fortune 500 company as a Unix sys admin. One of my projects was to assist in bringing a new Oracle financials system on line. The data on this system was so sensitive that only the executive board was given access, and then only via SecurID cards from specific locations during specific time windows.
Nightly backup tapes were queued in a fireproof walk-in vault before going offsite at the end of each day. I happened to be strolling by the vault one day and noticed the backup tapes sitting there on a shelf in the vault, right next to the open vault door. I did some checking and found out that the vault was left open during business hours so that the operations folks had easy access to backup media. The vault was in a different department than I/S, on a main hallway, right near the front door of the building. Obviously, I mentioned this to the Operations Manager. The new policy limited access to only a couple of operations supervisors, and instituted a media checkout policy (nothing a little social engineering couldn't thwart, but far better than the previous situation).
So what's the moral of the story? Make sure your security policy deals with backup media. Don't just assume that your operations department (or the offsite storage provider) is securely managing your media.
I know the author of a similar book that hasn't quite finished up yet. He was concentrating on the SAN's aspect of it since NAS security is pretty much the same FAQ as 'how to setup a file server'.
Secure SANs was slated to come out last year but hasn't ever been more then a link on Amazon. It dealt with the ugliness of iSCSI and how the 'air gap' security that protected this data for so long is now gone and storage administrators are struggling to learn how the real world works.
Not to bash storage admins but they've relegated most of their 'security efforts' to LUN masking and other such techniques. Now that SCSI drive commands are traversing networks huge security vulnerabilities are opened up. I read an advanced chapter of the Secure SAN title and the best part was an executive from a prominent NAS company stating that he wasn't worried about the security of the products since "only a handful of ppl in the world could have this conversation".
Check out the recent efforts at Storage Networking Industry Association who have come as close to working miracles as I've personally seen. They have managed to create some various technial frameworks and security processes that make vendors work together.
One interesting note about the book featured here is that it also deals with NAS and DAS. NAS and SANS have been fighting it out as IDE and SCSI have. One is cheap and easy the other pricey and very difficult. DAS on the other hand is a joke to me. The ability for one computer to change bits in another's memory DIRECTLY does not sound like a good idea. Hackers have worked for decades to write shell code that allows the ability to change bits in memory and now the storage industry has created a way to get directly in there bypassing all OS security.... yea great idea