Adaptive granule generation for parallel queries with run-time data pruning

A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform a method comprising:

• generating an execution plan for a query to be executed by a plurality of slave processes, said execution plan comprising a plurality of plan operators;
• for a first plan operator of said plan operators, generating a first set of work granules, and for a second plan operator of said plan operators, generating a second set of work granules assigned a plurality of partitions, each work granule of said second set of work granules being assigned to one or more of said plurality of partitions;
• assigning to a first subset of slave processes of said plurality of slave processes said first set of work granules;
• based on execution of said first set of work granules by said first subset of slave processes, making a determination that one or more partitions include no rows;
• based on said determination, modifying the second set of work granules to generate a modified second set of work granules;
• assigning at least some of said modified second set of work granules to a second subset of slave processes of said plurality of slave processes, said second subset of slave processes executing said at least some of said modified second set of work granules.