If you're working at a scale where you accumulate a ton of data quickly and don't know what questions you want to ask, it gives you nigh-infinite scalability without nigh-infinite maintenance, so you can ask questions later.

The other use-case is akin to giving a really big house to a hoarder.

> Can anyone explain the rise in DBs backed by object storage?

I dare say it's a "when the only tool you have is a hammer" problem.

But it might also be the fact that object storage services are reliable, relatively cheap by cloud standards, and do not have a upper bound on resource availability.

I think the main attraction of object storage like S3 is that it already solves the hard problems (distributed-ness, scalability, durability, etc) with a simple(r) API.

Also, apis like S3 are defacto standard with many vendors supporting a descent subset of it.

One other reason that hasn’t been mentioned is that it eliminates inter-AZ network fees. WarpStream has some numbers on their site that show huge savings by using S3 as the backbone.