[vincentdm]

I really like this addition. We store a lot of data for different customers, and most of our queries are only about data from a single customer. If I understand it correctly, if we would partition by customer_id, once the query planner is able to take advantage of this new feature, it will be much faster to do such queries as it won't have to wade through rows of data from other customers.

Another common use case is that we want to know an average number for all/some customers. To do this, we run a subquery grouped by customer, and then calculate the average in a surrounding query. I hope that the query builder wil eventually become smart enough to use the GROUP BY clause to distribute this subquery to the different partitions.