Partitioning and replication.

a. Explain why range-partitioning gives better control on tablet sizes than hash partitioning. List an analogy between this case and the case of B+-tree indices versus hash indices.

b. Some systems first perform hashing on the key, and then use range partitioning on the hash values. What could be a motivation for this choice,
and what are its drawbacks as compared to performing range partition direction on the key?

c. It is possible to horizontally partition data, and then perform vertical partitioning locally at each node. It is also possible to do the converse, where vertical partitioning is done first, and then each partition is then horizontally partitioned independently. What are are the benefits of the first option over the second one?