Inter-operation parallelism occurs when two or more operations are executed in parallel, either as a dataflow or independently.
One designates as dataflow the form of parallelism induced by pipelining. Independent parallelism occurs when operations are executed at the same time or in arbitrary order.
Independent parallelism is possible only when the operations do not involve the same data. To speed up processing of a query by executing in parallel the different operations in a query expression.
Parallel query processing of database clusters is based on two techniques known as infra-query and inter-query parallelism.
In intra-query parallelism, a query is decomposed in sub-queries that scan different subsets of the data. The sub-queries are executed inparallel in the cluster nodes.
