Jared Kim’s Weblog

Backup files that are on S3?

Posted by: Jared Kim on: July 28, 2006

I’m transitioning over a project to use S3 for file storage. However, one concern I have is the safety of the files.

Not that I think Amazon’s servers would crash, but what if someone discovered/cracked/haxored our ID & secret key and simply deleted everything in our bucket?

I’d feel much safer to have an easy way to keep a total backup of a bucket somewhere. Any ideas out there?/

3 Responses to "Backup files that are on S3?"

1 | cardbox

August 13th, 2006 at 4:45 pm

What is the size of the problem? (In gigabytes or terabytes)

2 | Jared

August 13th, 2006 at 6:06 pm

I’m estimating it might be in the hundreds of gigabytes

3 | cardbox

August 14th, 2006 at 8:35 am

Well, the requirements are less stringent for this sort of backup, because it doesn’t really need to be accessible till disaster strikes. Get a PC with enough storage and just download everything to it?

You might also just set up a second Amazon account with a different username, and have a regular process that copies objects across from your main account. You might reckon that that second username (not being used anywhere else) could be safer against theft.

If you’re interested in the risks as well as the benefits of using Amazon S3, I’ve started a blog series at http://cardbox.wordpress.com/2006/08/14/s3-in-business-1/ which will cover all the risks I can think of. It’s important to be a pessimist at the beginning of a project: it makes it likelier that the worst won’t happen, or that you’re prepared for it if it does!

Comment Form