(Warning: this post is for Mac users only. For everyone else, go look at the exploding sperm whale.)
My wife recently sold her old MacBook Pro to a student, and in one of those moments of profound embarrassment, the internal hard drive died not two weeks later. It wasn’t the original drive. In a quest for more space and a faster drive, my wife had swapped out the original 5400 rpm drive with a larger 7200 rpm version. It ran hot and loud, and we suspect it was the heat that killed it. There’s a reason Apple only sold that laptop with a 5400 rpm drive.
Since the local repair shop was going to charge this student an eye-watering €150 to replace the drive — which is more than we sold the laptop for — we offered to pay for half the cost of a new drive and install it ourselves. Total cost: €60. The new, 5200 rpm drive runs cool and quiet, and the student reports that the laptop’s fans don’t come on anymore, either. So that ended well.
All except for the part where I watched my wife putting in the last screws and asked, “What about her data? She had a backup, right?”
“Of course not,” said my wife. At my look of horror, she added that the student didn’t have anything critical on it yet, so it was all right. Still — no backup?
For a paranoid triple-backup freak like me, this was akin to hearing about a skydiver who jumped out of a plane without a parachute. If you don’t back up your computer’s data, at some point or another you will lose it. Hard drives fail, period. The only question is how long it takes. That’s why I have two separate bootable backups on external drives, and rotate them regularly. I also back up to the cloud, in case the external drives are simultaneously destroyed.
My free storage on Dropbox holds the smaller, critical stuff, like all of my writing, some of the contents of my home folder, and various software databases/preferences. But the big stuff can’t go on there — my photo library alone is now over 50 GB. Dropbox of course allows a paid upgrade for more storage, but it starts at $10/month, and I don’t want to pay that much.
Until recently, I used Amazon’s S3 cloud storage for my photo library, which at ten cents per gigabyte cost me five euros a month. I was happy with that until I learned about Amazon’s new service, Glacier. It’s one penny per gigabyte.
Glacier differs from S3 in that it’s meant for storage you don’t want to get to. The servers are low-energy, designed not to run constantly but rather to rest when not under active usage. The benefit is they cost much less to run, and the drawback is they’re not always running. That means you can’t download your data whenever you want. What’s more, the download speed is quite slow. If you want to download your data faster, then you pay a higher fee for it. And if you want to download all of your data, you pay a higher fee for that, too.
That’s fine with me — this is, after all, my “in case of disaster” backup. If I need that data, it means I’ve survived a house fire, a tsunami, or an earthquake. Maybe all three. In that event, I’m unlikely to care that it takes a few days and a few euros to download my entire photo library.
Since Glacier is ten times cheaper than S3, I’m now taking advantage of the cost to back up much more stuff: my iTunes library, a bunch of videos, some old Keynote presentations that I want to keep forever, my collection of wallpapers, and my mail archive. I’d always been a bit uncomfortable that not all of my digital life was protected by an offsite backup, but that’s no longer an issue. And I’m still paying less than I did with a much smaller backup on S3.
“Sounds great!” you say. “What’s the catch?”
The catch is that the user interface on Amazon’s Web Services is terrible. I have no idea how a company that turned online shopping into the simplest point-and-click endeavor on the planet could produce such a baffling swamp in its other branch. Believe me when I say that you don’t want to mess with AWS any more than you have to. That means setting up your account, choosing your region, and then running away.
Here is where Arq comes in. Once you have your Amazon account set up and the secret key codes in hand, you can fire up Arq, enter your key codes, and then go straight to work. Arq makes it easy. When you’re presented with the working screen, there’s really only one button to press, which says, “Add a Folder to Backups…” Once you’ve chosen your folder, you’re asked whether you want to use S3, or Glacier + S3. Arq asks this on every folder, so you can easily store some files on Glacier, and files that require faster access on S3.
One note: because of the way Glacier is designed, Arq must keep some metadata on S3 for every Glacier backup (which is why the option is “Glacier + S3”). Otherwise it would not be able to communicate in any reasonable length of time with Amazon’s servers. However, these metadata files are tiny and incur negligible expense.
Arq also gives you an easy way to exclude folders and files from the backup. In the screenshot above, you can see that the iTunes backup is excluding the folder iTunes Media/Mobile Applications — that’s because there’s no need for me to store my iPad apps on Glacier. I also excluded podcasts, home videos and audiobooks.
But ease of use isn’t even the main reason to use Arq. That distinction goes to its hard-working little sub-program called Arq Agent. You can tell Arq Agent to start itself automatically when you log in, after which it stays in the background, quietly monitoring your computer for any changes to the files you’ve selected for backup. When you make a change, Arq Agent prepares it for synchronization with the online backup. That synchronization takes place whenever you schedule it — for me, it’s every day at 23:00. If your computer isn’t awake during the scheduled backup, Arq Agent waits until it is. Once you’ve selected your files and set your schedule, you can forget all about it. Arq Agent takes care of all your backup duties for you, diligently keeping your files synchronized.
When you want to retrieve files from your online backup, just find the most recent backup in Arq and click the only button on the screen: Restore. And that part about recent backups? Arq does versioning, so if you’re using S3, it’s basically an online Time Machine. To keep those versions from bloating up into a big expense, you can set a budget and tell Arq to enforce it.
For me, Arq is indispensable. It’s also not cheap at $40, but how much does an insurance policy cost? At any rate, you can try it for free for 30 days. The trial download and more information are available at Haystack Software.