I was trying to convey my backup workflow/setup to @joeday in 140 and it just wasn’t working very well. Twitter – as one might expect – is not exactly the place for detailed technical discussions, but it does provide fertile ground to spark ideas and dialogue. I told @geekshui that I’d blog my setup and that turned out to be just enough of a catalyst to force me to iron out my strategy for rud.is and future (if any) non-cooking/family blogging.
Background
I’m [still] a die-hard OS X user, despite the increasing gatekeeper motif Apple is sporting these days. My main computer is a MacBook Pro which I would stupidly run back into a burning building to rescue. Everything is on it. Everything. I digitize receipts, house our multimedia, spin out VMs like a DJ, create, compose, torrent, rip, zip and hack from it. Consequently, ensuring my data is available is kinda important to me.
I’ve been around computers long enough to have learned some painful lessons from four simple characters: MTBF. Drives break. Electronics fail. It’s an undeniable fact. The only way to recover from these failures is to have a good strategy for keeping your data available.
Strategy #1: Backups
While hard to digest on Twitter, my backup strategy is pretty straightforward. I use Time Machine for OS-managed full system backups. I rotate these between two large (1TB & 2TB) hard drives and I retire one large hard drive each year (MTBF…remember?). This gets me individual file recovery pretty quickly over a decent time period and a bit of hardware piece of mind.
I also have two 2.5″ IEEE 1394 drives that I SuperDupe/CarbonCopyClone images to every time Apple issues a 10.x.y update. Again, I rotate between since I really don’t trust drive manufacturers. I haven’t relied on TrueCrypt for a while (which would make for an ugly workflow) for system volumes, but it’s easy to clone disks that have FileVault protected data as long as you do so from an account that does not use or rely on FileVault data.
Both Time Machine and the drive cloning can occur while I’m sleeping, so no workflow is impacted.
Strategy #2: Dropbox
I have to start by sharing just how much I <3 Dropbox. I don’t use the free service as I grew weary of keeping within the paultry limits. Getting a paid sub to it provides more than just freedom from minutiae. I now get (as long as they have no hiccups) full recovery back as far as I want in the event I do actually lose a file or two. I have Dropbox configured on my MacBook Pro, a home Windows machine and a home Linux box. This means that even if I lose the drive on my Mac, I can get some of my non-sensitive data back from one of the other Dropbox-enabled systems (which is much faster than recovering from backups). It also means that I can get right back to work on a different system – as long as I have not used an OS X-specific program.
I could rant for quite a while about Dropbox, but it should be pretty obvious why this is part of of my backup strategy.
Strategy #3: rsync.net
While Dropbox houses non-sensitive data offsite (again, assuming no service hiccups), there is a subset of my information that I do want housed off-site in the event there is a catastrophic issue with our abode. For that, I have been using rsync.net since it’s inception. They provide outstanding customer support, have a unique view and practices around warrants and fully understand the needs of technical users concerned about availability and privacy.
There are some other things we do to ensure a refresh of the content on media drives that get hooked up to our PS3 or displays, but the the above three steps are how I ensure that I always have access to the data that enables my workflow.
