Query plans for analytic SQL constructs
A system and method for managing data storage and data access with querying data in a distributed system without buffering the results on intermediate operations in disk storage.
1. A non-transitory computer-readable medium storing instructions which, when executed by one or more third processors of a computing device, cause the computing device to:
access an execution plan for a set of data referenced by a query;
initiate, within a first processor, a first parallel execution process in the execution plan to process the set of data and generate an intermediate result of a query;
push, by the first processor, during execution of the query plan, the intermediate result to one or more second processors in the query plan for processing by the plurality of secondary parallel execution processes within the plurality of second processors that generate a plurality of second results, wherein each of the one or more second processors store at least some of the plurality of intermediate results in a local cache that corresponds to that second processor;
and
store a final result to the disk storage within a storage platform.
2. The non-transitory computer-readable medium of claim 1 , wherein each of the first processor and the one or more second processors are decoupled from the disk storage.
3. The non-transitory computer-readable medium of claim 1 , wherein the plurality of second results is generated without writing the intermediate result to the disk storage.
4. The non-transitory computer-readable medium of claim 1 , wherein each of the plurality of secondary parallel execution processes the intermediate result with a different operation.
5. The non-transitory computer-readable medium of claim 1 , wherein the instructions further cause the computing device to:
delay operation of at least one of the plurality of secondary parallel execution processes so as to coordinate timing among other secondary parallel execution processes of the plurality of secondary parallel execution processes.
6. The non-transitory computer-readable medium of claim 1 , wherein the intermediate result comprises a plurality of rows of database data.
7. The non-transitory computer-readable medium of claim 1 , wherein each of the plurality of secondary parallel execution processes are unique operators.
8. The non-transitory computer-readable medium of claim 1 , wherein the intermediate result is not materialized.
9. The non-transitory computer-readable medium of claim 1 , wherein the intermediate result generated by the first parallel execution processes is not materialized to a temporary structure.
10. The non-transitory computer-readable medium of claim 1 , wherein the set of data is associated with a database system that includes a relational database.
11. The non-transitory computer-readable medium of claim 1 , wherein the set of data is associated with a database system that is a multi-tenant database that isolates computing resources and data between different customers.
12. The non-transitory computer-readable medium of claim 1 , wherein the each of first and the one or more second processors is coupled to the disk storage via a communications network.
13. The non-transitory computer-readable medium of claim 1 , wherein the instructions further cause the computing device to:
receive the query for information stored in one or more databases.
14. A method comprising:
accessing an execution plan for a set of data referenced by a query;
initiating, within a first processor, a first parallel execution process in the execution plan to process the set of data and generate an intermediate result of a query;
pushing, by the first processor, during execution of the query plan, the intermediate result to one or more second processors in the query plan for processing by the plurality of secondary parallel execution processes within the one or more second processors that generate a plurality of second results, wherein each of the one or more second processors store at least some of the plurality of intermediate results in a local cache that corresponds to that second processor;
and
storing a final result to the disk storage within a storage platform.
15. The method of claim 14 , wherein each of the first processor and the one or more second processors are decoupled from the disk storage.
16. The method of claim 14 , wherein the plurality of second results generated without writing the intermediate result to the disk storage.
17. The method of claim 14 , wherein each of the plurality of secondary parallel execution processes the intermediate result with a different operation.
18. The method of claim 14 , further comprising:
delaying operation of at least one of the plurality of secondary parallel execution processes so as to coordinate timing among other secondary parallel execution processes of the plurality of secondary parallel execution processes.
19. The method of claim 14 , wherein the intermediate result comprises a plurality of rows of database data.
20. The method of claim 14 , wherein each of the plurality of secondary parallel execution processes are unique operators.
21. The method of claim 14 , wherein the intermediate result is not materialized.
22. The method of claim 14 , wherein the intermediate result generated by the first parallel execution processes is not materialized to a temporary structure.
23. The method of claim 14 , wherein the set of data is associated with a database system that includes a relational database.
24. The method of claim 14 , wherein the set of data is associated with a database system that includes a multi-tenant database that isolates computing resources and data between different customers.
25. The method of claim 14 , wherein the each of the first and the one or more second processors is coupled to the disk storage via a communications network.
26. The method of claim 14 , wherein the instructions further cause the computing device to:
receive the query for information stored in one or more databases.
27. A system comprising:
a query coordinating process programmed to:
access an execution plan for a set of data referenced by a query;
a first processor programmed to:
initiate a first parallel execution process in the execution plan to process the set of data and generate an intermediate result of a query, and
push, by the one or more processors, during execution of the query plan, the intermediate result to the one or more second processors in the query plan for concurrent processing by the plurality of secondary parallel execution processes; and
the one or more second processors programmed to generate the plurality of intermediate results, wherein each of the one or more second processors store at least some of the plurality of intermediate results in a local cache that corresponds to that second processor; and
a storage platform programmed to store a final result to the disk storage.
28. The system of claim 27 , wherein each of the first processor and the one or more second processors are decoupled from the disk storage.
29. The system of claim 27 , wherein the plurality of second results is generated without writing the intermediate result to the disk storage.
30. The system of claim 27 , wherein each of the plurality of secondary parallel execution processes process the intermediate result with a different operation.