|< < 24 > >|

Unique

UniqueHash

If the input does not fit in memory

next(): Eliminate duplicates in each partition (one at a time, on demand):

  • Read the partition into memory.
  • Use an in-memory hash table (as when the entire table fit in memory).
  • Return rows out of the hash table.
  • When all rows returned, go on to the next partition.

|< < 24 > >|