Granted Patent B2

US 12,339,882 · App. 18/381,165 · Granted Jun 24, 2025

Inventors: Thomas Neumann (Munich, DE); Viktor Leis (Munich, DE); Alfons Kemper (Munich, DE); Tobias Muehlbauer (Munich, DE); Jan Finis (Munich, DE)

Assignee: Tableau Software, Inc.

G06F16/3332G06F16/23G06F16/334

Abstract

Claims (83)

1. A method for retrieving data from a database, comprising:

at a computer system having one or more processors and memory storing one or more programs configured for execution by the one or more processors:

receiving a database query from a client;

translating the database query into an intermediate representation;

computing an estimated query execution time according to an estimated number of rows that will be accessed to retrieve a result set corresponding to the database query;

computing an estimated compilation time to compile the intermediate representation into machine executable code according to characteristics of the intermediate representation;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a first optimized compilation criterion:

performing one or more optimization passes of a first type on the intermediate representation to form an optimized intermediate representation of a first type;

compiling the optimized intermediate representation of the first type to form optimized executable machine code of a first type; and

executing the optimized executable machine code of the first type to retrieve the result set from the database;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a second optimized compilation criterion:

performing one or more optimization passes of a second type on the intermediate representation to form an optimized intermediate representation of a second type;

compiling the optimized intermediate representation of the second type to form optimized executable machine code of a second type, wherein the optimized executable machine code of the second type has a lower execution time than the optimized executable machine code of the first type; and

executing the optimized executable machine code of the second type to retrieve the result set from the database; and

returning the result set.

2. The method of claim 1 , wherein computing the estimated compilation time is based on at least one of:

a number of instructions in the intermediate representation;

types of instructions in the intermediate representation; and

a number of functions in the intermediate representation.

3. The method of claim 1 , wherein computing the estimated compilation time is based on a number of execution blocks in the intermediate representation, wherein each block comprises a maximal contiguous sequence of instructions without a jump instruction.

4. The method of claim 1 , wherein:

the first optimized compilation criterion is satisfied when the estimated number of rows that will be accessed to retrieve the result set is greater than a first threshold but less than a second threshold; and

the second optimized compilation criterion is satisfied when the estimated number of rows that will be accessed to retrieve the result set is greater than the second threshold.

5. The method of claim 1 , further comprising computing a ratio of the estimated query execution time and the estimated compilation time, wherein:

the first optimized compilation criterion comprises a rule that the ratio is greater than a first threshold but less than a second threshold; and

the second optimized compilation criterion comprises a rule that the ratio is greater than the second threshold.

6. The method of claim 1 , wherein the first optimized compilation criterion and the second optimized compilation criterion partition a set of all pairs (e, c) of estimated query execution time, that are greater than a minimum estimated query execution time, and estimated compilation time into two disjoint regions.

7. The method of claim 1 , wherein in accordance with a determination that the database query includes a join operation for a first dataset and a second dataset, the estimated query execution time is adjusted by summing a first cardinality of the first data set and a second cardinality of the second dataset.

8. The method of claim 1 , further comprising:

in accordance with a determination that the optimized executable machine code of the first type is executed:

determining a compilation time for the optimized executable machine code of the first type; and

in accordance with a determination that the compilation time for the optimized executable machine code of the first type and the estimated compilation time are different, adjusting the estimated compilation time based on the compilation time for the optimized executable machine code of the first type.

9. The method of claim 8 , wherein, in accordance with a determination that the estimated compilation time has been adjusted, adjusting the first optimized compilation criterion and/or the second optimized compilation criterion.

10. The method of claim 1 , further comprising:

in accordance with a determination that the optimized executable machine code of the first type is executed:

determining a query execution time for the optimized executable machine code of the first type; and

in accordance with a determination that the query execution time for the optimized executable machine code of the first type and the estimated query execution time are different, adjusting the estimated query execution time based on the query execution time for the optimized executable machine code of the first type.

11. The method of claim 10 , wherein, in accordance with a determination that the estimated query execution time has been adjusted, adjusting the first optimized compilation criterion and/or the second optimized compilation criterion.

12. The method of claim 1 , further comprising:

in accordance with a determination that the optimized executable machine code of the second type is executed:

determining a compilation time for the optimized executable machine code of the second type; and

in accordance with a determination that the compilation time for the optimized executable machine code of the second type and the estimated compilation time are different, adjusting the estimated compilation time based on the compilation time for the optimized executable machine code of the second type.

13. The method of claim 12 , wherein, in accordance with a determination that the estimated compilation time has been adjusted, adjusting the first optimized compilation criterion and/or the second optimized compilation criterion.

14. The method of claim 1 , further comprising:

in accordance with a determination that the optimized executable machine code of the second type is executed:

determining a query execution time for the optimized executable machine code of the second type; and

in accordance with a determination that the query execution time for the optimized executable machine code of the second type and the estimated query execution time are different, adjusting the estimated query execution time based on the query execution time for the optimized executable machine code of the second type.

15. The method of claim 14 , wherein, in accordance with a determination that the estimated query execution time has been adjusted, adjusting the first optimized compilation criterion and/or the second optimized compilation criterion.

16. A computer system having one or more computing devices, each computing device having one or more processors and memory, wherein the memory stores one or more programs configured for execution by the one or more processors, the one or more programs comprising instructions for:

receiving a database query from a client;

translating the database query into an intermediate representation;

computing an estimated query execution time according to an estimated number of rows that will be accessed to retrieve a result set corresponding to the database query;

computing an estimated compilation time to compile the intermediate representation into machine executable code according to characteristics of the intermediate representation;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a first optimized compilation criterion:

performing one or more optimization passes of a first type on the intermediate representation to form an optimized intermediate representation of a first type;

compiling the optimized intermediate representation of the first type to form optimized executable machine code of a first type; and

executing the optimized executable machine code of the first type to retrieve the result set from a database;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a second optimized compilation criterion:

performing one or more optimization passes of a second type on the intermediate representation to form an optimized intermediate representation of a second type;

compiling the optimized intermediate representation of the second type to form optimized executable machine code of a second type, wherein the optimized executable machine code of the second type has a lower execution time than the optimized executable machine code of the first type; and

executing the optimized executable machine code of the second type to retrieve the result set from the database; and

returning the result set.

17. The computer system of claim 16 , wherein computing the estimated compilation time is based on a number of execution blocks in the intermediate representation, wherein each block comprises a maximal contiguous sequence of instructions without a jump instruction.

18. The computer system of claim 16 , wherein:

the first optimized compilation criterion is satisfied when the estimated number of rows that will be accessed to retrieve the result set is greater than a first threshold but less than a second threshold; and

the second optimized compilation criterion is satisfied when the estimated number of rows that will be accessed to retrieve the result set is greater than the second threshold.

19. The computer system of claim 16 , the one or more programs further comprising instructions for computing a ratio of the estimated query execution time and the estimated compilation time, wherein:

the first optimized compilation criterion comprises a rule that the ratio is greater than a first threshold but less than a second threshold; and

the second optimized compilation criterion comprises a rule that the ratio is greater than the second threshold.

20. A non-transitory computer readable storage medium storing one or more programs configured for execution by a computer system having one or more processors and memory, the one or more programs comprising instructions for:

receiving a database query from a client;

translating the database query into an intermediate representation;

computing an estimated query execution time according to an estimated number of rows that will be accessed to retrieve a result set corresponding to the database query;

computing an estimated compilation time to compile the intermediate representation into machine executable code according to characteristics a size of the intermediate representation;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a first optimized compilation criterion:

performing one or more optimization passes of a first type on the intermediate representation to form an optimized intermediate representation of a first type;

compiling the optimized intermediate representation of the first type to form optimized executable machine code of a first type; and

executing the optimized executable machine code of the first type to retrieve the result set from a database;

in accordance with a determination that the estimated query execution time and the estimated compilation time satisfy a second optimized compilation criterion:

performing one or more optimization passes of a second type on the intermediate representation to form an optimized intermediate representation of a second type;

compiling the optimized intermediate representation of the second type to form optimized executable machine code of the second type, wherein the optimized executable machine code of the second type has a lower execution time than the optimized executable machine code of the first type; and

executing the optimized executable machine code of the second type to retrieve the result set from the database; and

returning the result set.

Assignments (2)

Continuity (6)

References Cited (68)