Interpartition parallelism

Interpartition parallelism refers to the process of dividing a single query into multiple parts that run in parallel on different partitions of a partitioned database.

Interpartition parallelism can be categorized into two parts: data fetching and data processing. In terms of interpartition parallelism the data processing is always executed on multiple partitions to achieve the maximal performance. But regarding to data fetching, there are two scenarios, one is serial fetching and the other is parallel fetching. The serial fetching means all the data from remote data source are fetched from coordinator partition while parallel fetching means each partition fetches partial of the data from remote data source and combine the data fetched by each partition equals to the final result. In serial fetching scenario, the data will be distributed to other partitions from coordinator partition.