Getpage-workload based index optimizer

• US 20030167255A1
• Filed: 10/18/2002
• Published: 09/04/2003
• Est. Priority Date: 03/01/2002
• Status: Active Grant

• Current Assignee: Software Engineering GmbH
• Original Assignee: Software Engineering GmbH
• Inventors: GraBhoff, Hans-Peter, Klosterhalfen, Jorg, Breuer, Guido
• Granted Patent: US 7,047,231 B2
• Time in Patent Office: Days
• US Class Current: 707/1
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

The index optimization (XOP) program is a method to recommend new or improved performance indexes for use by an IBM mainframe DB2 database server for a given set of SQL statements. The recommendations are calculated based on a single call of the DB2 optimizer for each SQL statement with the existing index set, parsing of the SQL statements, the filter factors of the SQL statement predicates, and the getpage workload of an SQL statement. The getpage workload of an SQL statement is the read/change activity on each table and each index currently used for all invocations of an SQL statement. The XOP program computes new and improved performance indexes based on the getpage workload of all relevant SQL statements and does not use the DB2 optimizer to determine the usefulness of potential indexes.

Claims

1. A method for selecting optimized indexes for a relational database, comprising:

(a) collecting data about all database queries in a representative time interval;

(b) determining database tables which have been queried during the representative time period so as to have triggered activity above a predetermined amount;

(c) computing an optimized index set based on table read/change activity; and

(d) recommending the optimized index set to a user.

2. The method of claim 1, wherein the data about the database queries is determined by running a trace on all database queries in said representative time interval, the trace includes determining the frequency of execution and the read/change activity on each table and each index for each query traced.

3. The method of claim 1, wherein the queries comprise SQL queries.

4. The method of claim 1, wherein the read/change activity comprises getpage workload activity.

5. The method of claim 1, wherein the optimized index set is determined by computing the potential savings in read activity of potential index column combination for all query invocations on each table having a read activity above a predetermined level.

6. The method of claim 1, wherein tables with a high change activity are not considered for index recommendation if the potential index overhead exceeds the benefits in table read activity savings.

7. The method of claim 2, wherein the potential indexes are computed for each such table including the following factors:

(a) the table read/change activity of all queries accessing this table;

(b) the filter factors of the predicates of all queries accessing this table;

(c) the predicate properties of being indexable for all queries accessing this table; and

(d) the query access paths of all queries accessing this table.

8. The method of claim 1, wherein indexes are expanded for index only access to eliminate table read activity.

9. The method of claim 1, wherein the database comprises a DB2 database.

10. An apparatus for determining an optimized index set for a database, comprising:

(a) one or more processors; and

(b) a memory coupled to the one or more processors and containing one or more sequences of one or more instructions which, when executed by the one or more processors, cause the one or more processors to perform the steps of;

(i) collecting data about all database queries in a representative time interval;

(ii) determining database tables which have been queried during the representative time period so as to have triggered activity above a predetermined amount;

(iii) computing an optimized index set based on table read/change activity; and

(iv) recommending the optimized index set to a user.

11. The apparatus of claim 10, wherein the data about the database queries is determined by running a trace on all database queries in said representative time interval, the trace includes determining the frequency of execution and the read/change activity on each table and each index for each query traced.

12. The apparatus of claim 10, wherein the queries comprise SQL queries.

13. The apparatus of claim 10, wherein the read/change activity comprises getpage workload activity.

14. The apparatus of claim 10, wherein the optimized index set is determined by computing the potential savings in read activity of potential index column combination for all query invocations on each table having a read activity above a predetermined level.

15. The apparatus of claim 10, wherein the potential indexes are computed for each such table including the following factors:

(e) the table read/change activity of all queries accessing this table;

(f) the filter factors of the predicates of all queries accessing this table;

(g) the predicate properties of being indexable for all queries accessing this table; and

(h) the query access paths of all queries accessing this table.

16. The apparatus of claim 10, wherein indexes are expanded for index only access to eliminate table read activity.

17. The apparatus of claim 10, wherein the database comprises a DB2 database

18. A computer-readable medium carrying one or more sequences of one or more instructions for determining, in a relational database management system, an optimized set of database indexes, the one or more sequences of one or more instructions including instructions which, when executed by one or more processors, cause the one or more processors to perform the steps of:

(i) collecting data about all database queries in a representative time interval;

(ii) determining database tables which have been queried during the representative time period so as to have triggered activity above a predetermined amount;

(iii) computing an optimized index set based on table read/change activity; and

(iv) recommending the optimized index set to a user.

19. The medium of claim 18, wherein the data about the database queries is determined by running a trace on all database queries in said representative time interval, the trace includes determining the frequency of execution and the read/change activity on each table and each index for each query traced.

20. The medium of claim 18, wherein the queries comprise SQL queries.

21. The medium of claim 18, wherein the read/change activity comprises getpage workload activity.

22. The medium of claim 18, wherein the optimized index set is determined by computing the potential savings in read activity of potential index column combination for all query invocations on each table having a read activity above a predetermined level.

23. The medium of claim 18, wherein the potential indexes are computed for each such table including the following factors:

(i) the table read/change activity of all queries accessing this table;

(j) the filter factors of the predicates of all queries accessing this table;

(k) the predicate properties of being indexable for all queries accessing this table; and

(l) the query access paths of all queries accessing this table.

24. The medium of claim 18, wherein indexes are expanded for index only access to eliminate table read activity.

25. The medium of claim 18, wherein the database comprises a DB2 database