Is it more efficient to use AWS s3 versioning to store gzipped mysql dumps?

Question

With my database, we are taking a mysql dump for backup every hour. We then upload that backup to AWS s3 for safety and redundancy. We are using versioning to only store the delta of each new backup. My question is, are we really gaining anything by gzipping the data first? Is the data going to be scrambled in gzip so that all the data will have to be re-stored with each version?

If we instead store an un-compressed version of the database will we actually save space over time due to more efficient versioning?

I should add that we have been storing the hourly gzipped backup in s3 for several months now using versioning. But I am unable to figure out in the online interface how to find how much each new version takes up in storage.

Answer 1

You seem to have S3 versioning and EBS snapshots confused. S3 versioning does not store a delta, it stores the entire version each time. Creating a backup where you only pay for the size of the delta is a feature of Elastic Block Store snapshots.