Typical Data Flow with 5 steps:
-
1 A client process issues a query
-
2 The statement is parsed and a shared server process executes the statement
-
3 The shared server process determines if the statement can utilize parallel processing, and then enlists two or more query servers to execute the statement in parallel. The shared server process partitions the work of each operation in the query equally among the query servers
-
4 Query servers process certain operations in the query independently and return sub-results back to the shared server process
-
5 The shared server process reassembles the results from the query servers and returns the query's results to the client
|
