Saturday, August 4, 2007

Jungle Disk Rules

I've decided I need to get better about backing up my computers. My buddy Cuong was telling me how much he loves his new personal NAS. (50 mbs with afp over wifi, wow!)

I've been trying to minimize my lifestyle by getting rid of physical things. I was getting all excited about buying one of these filers, and then I realized that buying another compute device goes against the zen minimalist philosophy that I've been striving for lately.

I remembered that Amazon has a cool storage service called S3. But it's a storage service, not a distributed file system or a backup solution. I figured someone must have written a utility that makes it possible treat S3 as a file system. I also thought there might be a nice Ruby mapping to S3. Turns out I was right on both counts.

There is a Ruby interface for S3 called AWS::S3. It's a nice wrapper on the S3 APIs. This is cool but not exactly what I'm looking for. I may want to use it for other projects down the road, so I'll keep it in mind. I just don't have time to write my own backup program on top of these APIs and the S3 service.

Luckily someone has already built an open source program called Jungle Disk. Jungle Disk will allow you to mount a network disk that is backed by the S3 service. After you configure it with your S3 credentials you're good to go. You can treat it like a normal network disk and copy files back and forth to it. It also has a built in backup feature. The backup feature is pretty basic, you can specify which files and folders to backup. You can set a schedule or simply run it manually. It's not as flexible as ChronoSync, but it's not bad. For instance, you can't setup different backup sets and schedule them at different intervals. If you really want this feature, you could use ChronoSync on top of Jungle Disk. I tried this at first, but it didn't seem to work well. Jungle Disk seemed to get wedged and I had to Force Quit the application.

I've been trying the built-in backup feature and that seems to work better. It handles my flaky Comcast internet connection gracefully. One problem I've seen is trying to backup my email folders while using email. This seems to cause problems for both Jungle Disk and Of course, this is a place where it would be nice to have separate backup sets. One that does my documents folder and another that does email. I could then run the email backup when I don't have mail running, but I could leave the document backup on autopilot.

Surprisingly this runs on Windows, Mac, and Linux. The code is open source so you can even check out their coding style. I'm using it on a Mac but I'll probably install it on my Windows boxes as well. Time to push all my storage into the cloud!

BTW, Jungle Disk is free for 30 days and then $20 to buy afterwards. I liked it enough to buy it. For your $20 you get lifetime upgrades and you can use it on as many computers as you like, as long as it's with the same S3 account. They also only accept Amazon Payments for the purchase. This makes me suspect that these guys are in a very close relationship with Amazon, and maybe even a part of Amazon. (Not that there is anything wrong with that.) It's just odd that you can't find out anything about the Jungle Tools LLC from their web site.

BTW, the Amazon Payments thing is very new too. (Read Scoble's take here.) I found it simple to use when paying for my Jungle Disk license. It seems the online payment space is getting interesting. Google Checkout and Amazon Payments seem to be going after Pay Pal with a vengeance. Gotta love competition!