SYSTEM AND METHOD FOR DISTRIBUTING QUERIES TO A GROUP OF DATABASES AND EXPEDITING DATA ACCESS

• US 20080104089A1
• Filed: 10/30/2007
• Published: 05/01/2008
• Est. Priority Date: 10/30/2006
• Status: Active Grant

• Current Assignee: Execue Incorporated
• Original Assignee: Execue Incorporated
• Inventors: DASARI, Viswanath, PATIL, Abhijit A., PRAGADA, Sreenivasa Rao
• Granted Patent: US 9,747,349 B2
• Time in Patent Office: Days
• US Class Current: 707/100
• CPC Class Codes: G06F 16/2471 Distributed queries
• CPC Class Codes: G06F 16/283 Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP

Abstract

Disclosed are systems, methods, and computer readable media for distributing queries to a group of databases and expediting data access. The method embodiment comprises receiving a query for information from one or more databases, and returning a result to the query for information requested in the query from a hybrid database, the hybrid database being constructed by the steps comprising: (1) identifying one or more source member databases where query acceleration is required, (2) identifying one or more common dimensions or grouping patterns in a plurality of databases, (3) creating one or more sets of commonly occurring common dimensions, (4) building one or more multi-dimensional databases using the one or more sets of commonly occurring common dimensions, (5) building one or more fractional databases, (6) compacting the one or more fractional databases into a master fractional database, (7) building a hybrid database metadata repository combining metadata from the master fractional database and the one or more multi-dimensional databases and source databases together referred to as derived member databases, and (8) building a query governor to interface between the query, the hybrid database metadata repository, the plurality of databases, the master fractional database, and the one or more multi-dimensional databases.

Claims

1. A method for database querying, the method comprising:

receiving a query for information from one or more source databases; and

returning a result to the query for information requested in the query from a hybrid database, the hybrid database being constructed by steps comprising;

(1) identifying one or more source member databases where query acceleration is required;

(2) identifying one or more common dimensions or grouping patterns in a plurality of databases;

(3) creating one or more sets of commonly occurring common dimensions;

(4) building one or more multi-dimensional databases using the one or more sets of commonly occurring common dimensions;

(5) building one or more fractional databases;

(6) compacting the one or more fractional databases into a master fractional database;

(7) building a hybrid database metadata repository combining metadata from the master fractional database and the one or more multi-dimensional databases and source member databases together referred to as derived member databases; and

(8) building a query governor to interface between the query, the hybrid database metadata repository, the plurality of databases, the master fractional database, and the one or more multi-dimensional databases.

2. The method of claim 1, wherein the source and derived member databases are either detailed record-level databases, aggregated cubes, fractional databases or any other structured content.

3. The method of claim 1, wherein the query governor processes queries input using common business terms to translate the input queries to queries specific to any of the source or derived member databases.

4. The method of claim 1, wherein identifying one or more common dimensions or grouping patterns is optimized for at least one parameter.

5. The method of claim 1, wherein compacting the one or more fractional databases when constructing the hybrid database is achieved by slice numbering each sample in a full database, and building a plurality of fractional databases by selecting samples starting from the same slice number to maximize record overlap across the plurality of fractional databases.

6. The method of claim 1, wherein the metadata of derived member databases is automatically absorbed into the metadata repository using database entity ID of source databases as the and base entity ID of derived member databases in the metadata repository.

7. The method of claim 1, wherein in the query governor optimizes query response time by at least one of (1) directing queries to faster, pre-computed and aggregated cubes, (2) sampling data marts or dynamically sourcing additional samples to meet required accuracy level or (3) querying source databases based on input query, required accuracy level and performance metrics of member databases.

8. The method of claim 1, wherein the query for information may include a desired level of accuracy of query results.

9. The method of claim 4, wherein the hybrid database is optimized based on a statistical approach to quantitatively determine grouping candidates or common dimensions based on historical or ongoing query usage or database distributions.

10. The method of claim 8, wherein the hybrid database system dynamically sources additional samples or slices to meet required accuracy level of query results.

11. The method of claim 2, wherein the source databases are incorporated into the hybrid database system as a member database by absorbing metadata into the metadata repository.

12. A system to query databases, the system comprising:

a module configured to receive a query for information from one or more source databases; and

a module configured to return a result to the query for information requested in the query from a hybrid database, the hybrid database being constructed by steps comprising;

(1) identifying one or more source member databases where query acceleration is required;

(2) identifying one or more common dimensions or grouping patterns in a plurality of databases;

(3) creating one or more sets of commonly occurring common dimensions;

(4) building one or more multi-dimensional databases using the one or more sets of commonly occurring common dimensions;

(5) building one or more fractional databases;

(6) compacting the one or more fractional databases into a master fractional database;

(7) building a hybrid database metadata repository combining metadata from the master fractional database and the one or more multi-dimensional databases and source member databases together referred to as derived member databases; and

(8) building a query governor to interface between the query, the hybrid database metadata repository, the plurality of databases, the master fractional database, and the one or more multi-dimensional databases.

13. The method of claim 12, wherein compacting the one or more fractional databases when constructing the hybrid database is achieved by slice numbering each sample in a full database, and building a plurality of fractional databases by selecting samples starting from the same slice number.

14. The method of claim 12, wherein the hybrid database is constructed to optimize accuracy or speed.

15. The method of claim 14, wherein the hybrid database is optimized based on a statistical approach to quantitatively determine candidates for grouping based on one or more of usage, database distributions and common querying patterns.

16. The method of claim 12, wherein the query for information includes a desired level of accuracy.

17. A computer-readable medium containing instructions for database querying, the instructions comprising:

receiving a query for information from one or more source databases; and

returning a result to the query for information requested in the query from a hybrid database, the hybrid database being constructed by steps comprising;

(1) identifying one or more source member databases where query acceleration is required;

(2) identifying one or more common dimensions or grouping patterns in a plurality of databases;

(3) creating one or more sets of commonly occurring common dimensions;

(4) building one or more multi-dimensional databases using the one or more sets of commonly occurring common dimensions;

(5) building one or more fractional databases;

(6) compacting the one or more fractional databases into a master fractional database;

(7) building a hybrid database metadata repository combining metadata from the master fractional database and the one or more multi-dimensional databases and source member databases together referred to as derived member databases; and

(8) building a query governor to interface between the query, the hybrid database metadata repository, the plurality of databases, the master fractional database, and the one or more multi-dimensional databases.

18. The computer-readable medium of claim 17, wherein compacting the one or more fractional databases when constructing the hybrid database is achieved by slice numbering each sample in a full database, and building a plurality of fractional databases by selecting samples starting from the same slice number.

19. The computer-readable medium of claim 17, wherein the hybrid database is constructed to optimize accuracy or speed.

20. The computer-readable medium of claim 19, wherein the hybrid database is optimized based on a statistical approach to quantitatively determine candidates for grouping based on one or more of usage, database distributions and common querying patterns.

21. The computer-readable medium of claim 17, wherein the query for information includes a desired level of accuracy.