Skip to content

The join across machines

students is sharded by student_id, and colleges is a small table of 400 rows. A query joins them.

What is the cheapest arrangement?