Skip to content

Backing up with Amazon S3

Amazon S3 is Amazon’s cloud storage service. At .15 per gigabyte, 50GB will cost me $7.50 a month, versus $20 for Dropbox or UbuntuOne. And odds are those services are planting your data on Amazon servers anyway! It’s a damn good deal for redundant storage. Both of these services give you 2 GB for free, so if you just want to back up some documents and spreadsheets the free options should work well for you. There are a variety of tools that will streamline the process, Jungledisk probably being the most popular. However, Jungledisk does cost an extra $3 a month, and creates a cached copy of what you store, so in effect it’ll eat up a lot of local disk space unnecessarily, as the local data already lives on my media sever.

A GUI just seemed to add overhead I didn’t need to deal with.Then I found s3cmd, an open source command line tool for Linux. Win!!! It can do an Rsync with Amazon, but again, not really necessary. My MP3 library is not dynamic. I add on average 1 album a month, So I’m copying my music and photo library up to Amazon, one directory at a time, in the background as I do other stuff on the laptop. This also gives me an opportunity to decide if some of the music on my server is really worth keeping. Once I get the photos and music up, I probably will set up rsync with my docs directory. I already back up the media server to Breck’s PC, so the Amazon files are triple redundant.

However you do it, make sure you are backing up everything important.

{ 2 } Comments

  1. JP | September 2, 2010 at 7:02 am | Permalink

    JungleDisk is awful. I’ve yet to find something that works the way I want for off-site automated backup, but I used S3 Bucket Explorer for awhile and really liked it. That was before they had scheduling and automation, so they appear to have improved those areas (it used to be free, too).

  2. Charles Feduke | September 2, 2010 at 11:20 am | Permalink

    Hah, so funny, I was just looking at the RSS feed for my JungleDisk backup and then hopped to this post next.

    I use JungleDisk on my server at my store – works flawlessly – JungleDisk for WHS (no longer supported) on my home server – again flawlessly – and s3cmd with a shell script for my remote Linux server. I bought into JungleDisk when it was cheaper ($20 unlimited), and my monthly fee is $1 with unlimited installs.

    Though I am playing around with the idea of moving my dedicated remote Linux server to EC2 with S3 as the file system for volatile data, thus making my current dedicated box obsolete (okay, specs wise it already is). Also have an itch to mess around with LISP on a Linux install – been doing it on OS X for a tad now.