Backing Up to Amazon’s S3

I’ve been playing with Amazon’s S3 service the last few days, and it really is interesting.  Now that some applications are starting to be built on top of it, even though it’s a relatively low-level service it can actually be useful for end users.

The specific scenario I’ve found it useful for is backup.

I have about 10 gig of photos that I’ve taken over the years, and I’d hate to lose them.  10 gig isn’t that much these days, and finding space on my local disks isn’t hard, but for irreplaceable data like photos, having off-site backups is important to me.

I’ve tried various ways of managing this – burning DVDs and keeping some at work for example – but I generally forget to update them so they end up out of date, and really for something this critical, I don’t trust having the data on DVD’s.  If for some reason I need the data on the backup discs, it’d be just my luck to find out that one of them has decayed or for some reason isn’t readable.

S3 is raw backup.  I used a tool called s3backup to take my photos and dump them into S3, and now I have my offsite backup.

What makes S3 unique is the price.  It’s 15 cents per gigabyte per month.  My 10 gigabytes of photos works out to about $1.50/month, which is much cheaper than any other online backup solution I’ve looked at.

Why do I trust S3 where I don’t trust my own backups?  Good question.  Amazon’s business is preserving this data, so if they ever lose data (for anyone) it’ll be time to revisit this.  But for now, the $1.50/month it’s going to cost me to keep my data in S3 seems like money well spent in not having to manage my own offsite backup.

SmugMug would have worked for this as well – but it’s $39/year, and I don’t need the photo management services they offer.  If I ever get to storing enough data that $39/year is cheaper than what I’m paying Amazon, I’ll consider switching.