My OS X data backup system
I’ve finally joined one of the smallest clubs in the world — the “I back up my data reliably, redundantly and automatically” club. It’s a good membership to have. It offers peace of mind, and protection against accidental data deletion. Here’s how it works.
My local hard drive
The hard drive on my MacBook Pro is 160GB. My home directory, minus my “Downloads” folder, is about 70GB.
Step one: Time Machine
“Backup for the rest of us”
Time Machine is one of OS X Leopard’s most useful features. It’s hassle-free backup for non-geeks. I do an hourly snapshot of the entire hard drive (minus my “Downloads” folder and minus any cache directories). Although technically a Time Machine backup could be used to restore my data in the event of a hard drive failure, this is not what I use it for. I primarily use it so that I can revert to older versions of a file or “undelete” something that I didn’t mean to get rid of. This is because Time Machine backups are not bootable, and take a long time to restore. My Time Machine hard drive is 500GB and it is slightly more than half full.
Time Machine recommendations: exclude directories where big files are often changing, or your backups will churn. That’s why my “Downloads” directory is omitted. A lot of big files live there temporarily.
Step two: SuperDuper
“Heroic system recovery for mere mortals”
SuperDuper costs $27.95 and is worth every penny. It functions as a hard drive cloner. Every night at 5:30am, it clones my hard drive to an external hard drive of the same size (160GB). I have it set up to do a “smart update backup” which means that it only updates files that have changed (so it only takes about an hour). Why do I do this when I have an hourly Time Machine backup? Because it creates a bootable backup. Here is what would happen if my hard drive failed: I’d restart, hold down Option-Command-Shift-Delete as it booted, and then select the external SuperDuper backup drive as my boot device. That’d get me going with my system as it existed as of 5:30am that morning. I could then access TimeMachine backups to access the files I’ve been working on, and that’d get me back to within one hour. I’d be back up and working in about 5 minutes, and I’d only lose (at most) an hour of work. That gives me a lot of peace of mind.
Step three: JungleDisk
“My house burned down, now what?”
JungleDisk is my safety net in case my laptop hard drive and both my backup drives crash, are stolen or are destroyed. JungleDisk is an online backup tool that interfaces with Amazon’s S3 storage service. You pay $20.00 for the JungleDisk software, and then you pay Amazon $0.15 per GB-Month of storage used, and $0.10 per GB of data uploaded. I have JungleDisk back up my home directory (minus Downloads and minus cache directories — about 70GB) every night at 4am. I pay anywhere from $15 to $30 a month (to Amazon). Restoring from S3 would be a long, horrid task. I hope I never have to do it. This is strictly my safety net in case of a catastrophe. This is so that I don’t die running into my burning house in order to try to save my data.
So there are three components. All of them operate automatically. This is critical. A backup system that relies on human memory is no backup system at all.
How do you back up? If you don’t, what is stopping you?
