In hash join of R and S, we first partition each relation using a hash function h, generating partitions R1,,Rm and S1,,Sm. In the join phase, we join each pair (Ri,Si) using another hash function h2. Describe the method for handling the following two cases in order to minimize the IO cost.

1) One of Ri and Si is larger than the memory (the other is smaller than memory)

2) Both of Ri and Si is larger than the memory