Amazon is working on an interesting new feature for the S3 service: Object versioning.
Once you enable versioning for one of your S3 buckets, any time you change an object in that bucket a version of the prior object will be stored in addition to the latest one. You can then perform operations on prior object versions such as retrieving older data, restoring "deleted" objects, and generally maintaining a fail-safe history of everything that happens in the bucket.
This will be a boon to anyone who is worried about their S3 data being accidentally deleted or corrupted by user/computer error.
The feature is currently in early beta form and is available for testing with buckets located in the "us-west-1" location. You can read about the current functionality here: Versioning Beta Design.
Better still, you can grab the latest JetS3t code from CVS and try it out for yourself! The code samples file CodeSamples.java now includes a section called "Bucket Versioning (Beta)" to get you started.
Both the versioning feature itself and JetS3t's support for it are in an early stage so watch out for warts.
Comments