Adaptive Server considers parallel access methods for queries that return aggregate results in the same way it does for other queries. For queries that use the group by clause to return a grouped aggregate result, Adaptive Server also creates multiple worktables with clustered indexes—one worktable for each worker process that executes the query. Each worker process stores partial aggregate results in its designated worktable. As worker processes finish computing their partial results, they merge those results into a common worktable. After all worker processes have merged their partial results, the common worktable contains the final grouped aggregate result set for the query.
