Date Posted: October 9, 2008
Update: October 24, 2008 New version adds port number entry on harvesting screen and replaces TRIM() function (DB2 9 only) with its equivalent LTRIM(RTRIM()) (supported on DB2 8 and 9).
What is QueryScope?
QueryScope is a visualization technology that transforms a complex, multi-page SQL statement into a colorful bubble diagram that shows the most important characteristics of the query. The visual cues stand out to signify relative sizes of database tables, table join patterns, and index availability. When multiple SQL statements are displayed side by side, their commonality and differences become immediately apparent. QueryScope provides a significant improvement over text-based query analysis tools, which break down when used for understanding and comparing mult-page SQL statements.
How does it work?
QueryScope is implemented in Java™ and runs in Java Run-time Environment (JRE) 1.6 or above (it has been tested on Java Development Kit (JDK)/JRE 1.6). The tool loads workload information from structured XML, such as QueryScope_Demo.xml (included). Its harvest function retrieves table, index, and workload information from user-specified and system tables in IBM® DB2® UDB Linux®, UNIX®, and Windows® and transforms them into structured XML format.
About the technology author(s)
The inception of the QueryScope technology was when the authors participated in a series of large data warehouse projects in which database administrators had to learn multi-page queries generated by typical business intelligence tools. Tha authors felt that a better technique was needed than scouring through page after page of SQL statements; they realized that visual summarization highlighting similarities and differences among problem queries could better hint at applicable tuning techniques. The authors are also working to extend the technology to a knowledge base of query tuning.
The authors are from the Database Research group at IBM Thomas J. Watson Research Center.
