In the vertical partitioning section, the solution proposes storing user related tables separately from file/chunk tables, but poses the challenge of doing joins across them. Makes sense. But the subsequent options are focused solely on file - chunk relationships in range or hash based partitioning. This never answers how we’re managing joins across the partitions. More fundamentally, the data model isn’t even clear in what it is storing. I imagine it to involve a User table, User to file table, file to chunk table, but that’s my guess.
Would be nice if somebody/instructors could clarify.