This document describes the IBM® DB2® Content Manager Query Performance Guide.
This guide, document version 1.1 dated June 27, 2008, covers performance implications of the DB2 Content Manager Version 8 query component. It is written for customers, partners, application designers and developers, systems and database administrators, and IBM Level 2 and Level 3 technical support representatives who need to increase efficiency and resolve query performance issues.
The document describes the DB2 Content Manager query processing flow, common ways to analyze and address performance problems, database tuning steps such as updating statistics and creating indexes, and ways to narrow down where time is spent. It then lists common rewrites to DB2 Content Manager queries, including handling "child OR" queries, conditions on child component attributes, the IN operator, text search instead of LIKE, wildcards, case-insensitive search, upward traversal queries and folder queries. Later sections cover logging information needed by support teams, XML data model export, and references.
| Product version covered | DB2 Content Manager Version 8 |
|---|---|
| Document version and date | 1.1 (June 27, 2008) |
The query received by the CM Query API is translated into SQL using information from the Data Model API, then sent to the Search Stored Procedure where ACL conditions are added and the query is executed. The result is a list of IDs called PIDs used to create DDOs.
After creating a new index, run the reorg, runstats, and rebind commands so that the DB2 UDB optimizer can effectively use the new index.
Using a wildcard at the beginning of the search term may prevent the database system from effectively using the index.
It is primarily intended for DB2 Content Manager external and internal users, including customers, partners, application designers and developers, administrators, and IBM L2 and L3 support representatives.