Does anyone use versioning in s3 for backups?

Question

Amazon Web Services

Maybe there are already ready-made scripts for backups with versioning, in order not to manually think about incrementality and backup the site with one command?

Answer 1

duplicity.nongnu.org/index.html can use S3 as a backend. Haven't gotten around to trying it myself yet, but I've only heard good reviews.

Answer 2

The site is already stored on S3, or on a regular file system, and is it planned to keep only backup copies in S3?
If the latter, then most likely it will be really easier to use something like duplicity. A ready-made, sometimes even proven backup solution is better than something built on the knee in half a day :).
If the former, then versioning is convenient to use only for quick treatment of user errors (accidentally overwritten the wrong file), IMHO. For S3 has such a vile limitation as 1000 elements in the response to a request. That is, it will not work to get a list of objects with all their versions in one simple movement, so that later it can be parsed and built, say, a “snapshot” of the site as of a certain point in time, it will not work, you will have to go out of your way and take into account a sufficient number of edge cases. But yes - it will be possible to get a snapshot of the site files in general at an arbitrary point in time, and not once an hour / day, as with a traditional backup. But is the game worth the candle?