I guess it's much better suited for small data scenarios. Still have to test it, but I think there's much better tooling support for it.

Comments