I'm not sure how OLAP relates to CockroachDB. The use case for CockroachDB is quite simple. You have millions of applications that runs Postgres. A large chuck of them have to process more and more data which a single master cannot handle and there is no easy way of scaling. Any option you pick does require a lot of changes to the codebase. This is were CockroachDB should come in handy. The dream is to have a drop-in replacement for Postgres that would scale, so all SELECT .. JOIN still works.