General questions about snappydata

[singhals]

Hi

Firstly, snappydata is amazing! We were having issues with joining large datasets and having the min/max column statistics gave us an incredible boost. I had some follow up questions regarding two things:

1. Is it valuable to pre-sort my parquet set columns that I'm joining on hopefully organizing the column buffers to allow for a smaller number of mix/max ranges making querying and joining more efficient?

2. Is it possible to partition by two columns on two different tables AND co-locate the two tables? We have two columns on both tables: an id and a type and we want to evenly distribute those across nodes but it seems like we get an error trying to colocate.

Thanks!