Amazon S3 Does Version Control
Amazon have set up version control on their S3 system ; it is currently in Beta (isn’t everything?) but looks like it could be very handy.
Having a scan over it the main problem I see is that instead of storing diffs between the versions of file they are storing the entire file again as a new object. As storage on s3 is insanely cheap anyway, you wouldn’t think its that big a deal, but think about it for a second.
If you are editing a high res video and storing it on s3, you could see your bills sky rocket. Lets say the video is 1gb, and you need to make lots of amends and get client approval for each one. With normal version control (something like dropbox I think is a suitable comparison here) you’d only be saving the changes between the original file and the current one. With s3 it’s an entire new file.
That is where you could easily end up using 10gb of storage by just amending a single file a few times. I can understand why, storing entire copies is easier then trying to figure out a diff engine capable of handling obscene file sizes.
blog comments powered by Disqus
