Cryptosmith

This is the first of two postings that talk about using RAID with Mac OS X. The second post talks specifically about OS X. This post talks about RAID and how I use it.

Thanks to RAID, I’ve developed a fairly reliable way to store lots of data with minimal risk of losing it all. The RAID mechanism itself deals with the risk of hard drive crashes. I also keep a spare hard drive away from home to reduce the risk of losing everything.

In the Good Old Days, all the data I really wanted and needed to keep forever was on paper printed with ink. Ideally, the paper was low-acid content so it could sit on a shelf or in a fireproof box for decades.

Today, I have all sorts of data stored on my hard drive that I really and truly want to save forever:

• Family photographs
• Contact lists
• Personal library of digitized reference materials
• Tax records
• Music library
• A growing collection of digital video clips

I used to save old data on diskettes, then on CDs, and then on DVDs. I can no longer read the oldest diskettes, or even some of the slightly newer ones. I’ve never really had trouble with home-burnt CDs or DVDs (assuming they’re not stored in a bag mixed with sand and metal tools). According to reports, however, the home-burnt disks aren’t really stable enough to serve as archives. Moreover, it takes a lot of work to burn sets of disks and store them safely.

With today’s ballooning hard drive sizes and shrinking prices, I just keep everything on my main hard drive. If it’s all on the main drive, then I can always get to it, instead of having to search for a missing CD or DVD. Also, I’m much more likely to see if something’s wrong or missing on my main hard drive than in a drawer somewhere. Data seems safer somehow if I have it at my fingertips.

On the other hand, there are big risks:

• I could lose everything in a household disaster, like a fire or burglary.
• I could lose everything if the hard drive crashes. This is even more likely.

A third problem is that I could wipe out things by accident, or some wayward software could do it for me. There are ways to reduce the risk of that, but that’s a topic for a different discussion. For now let’s worry about “natural” disasters. That’s what RAID is for.

What is RAID?

Hardware breaks. Hard drives store incredible amounts of data packed into millionths of an inch. A tiny arm carries magnetic read/write heads that fly within millionths of an inch of the disk surface. If the arm bends too much, the head “crashes” into the spinning disk surface. This ruins the heads and leaves a nasty groove in the disk, making it impossible to retrieve much data. (Well, the right lab might be able to do it).

In practice, hard drives are very reliable. They tend to fail when they’re brand new, or after they’ve been in service for several years. Very few of us have actually suffered a drive failure of any kind, much less a head crash. But it happens, and drives will fail more often as competition pushes hard drive prices lower and lower.

RAID lets you take two or more cheap hard drives and “run them in parallel” to make them faster, more reliable, or both. Personally, I care about reliability, so I use “RAID 0″ which writes everything to two separate hard drives. If one dies, the other has a copy. The system keeps running with the copy.

This is also called “mirroring” since each drive is an exact “mirror” copy of the other. If I do lose one of the drives, I just throw out the broken one, get a new one, install it, and the system copies the working drive to the new one.

To make RAID work, your computer either needs RAID software installed in the operating system or built-in RAID controller hardware. I’ve used both. Either is fine as long as it really works.

What about “household disasters?”

By itself, RAID only deals with hard drive crashes. If you lose the whole computer system through fire, flood, theft, or some other disaster, you’ll lose the whole RAID set.

To defend against this, I actually keep 3 hard drives: two are running in my home machine and the third is at the office (it’s an “off site backup”). Every few weeks I swap out the extra drive at work and exchange it with the one at home. If I call the drives, A, B, and C, this works as follows:

• We start with Drives A and B in my home machine; Drive C at work
• I bring Drive C home.
• I turn off the computer and remove Drive B.
• I erase Drive C.
• I install Drive C in the computer and turn it on.
• I add Drive C to the RAID set, and the system copies A onto C.
• I take Drive B to work and leave it there for a few weeks.
• I repeat the process, swapping B and C again.

If something really nasty happens to my home machine, I bring Drive C home and I’m back up and running.

The devil is in the details

I’ve been using RAID on and off at home for 3 or 4 years. I’ll get into the details in another post, and provide some horror stories as well.

Posted under Household Tech |