Indexing IBM Content Cortex databases

Indexing Content Cortex databases can improve the performance of IBM Content Cortex.

Consider creating database indexes when you want to improve the performance of the Content Cortex database.

Reading documentation data tables

Documentation data tables that have information about database indexes are used throughout the performance tuning topics so that you can create performance enhancing database indexes on one (single) or more (multiple) database columns.

In documentation, these indexes are typically referred to as “single-property indexes” and “composite-properties indexes”.

Each table is introduced by a short summary that describes why the index is beneficial. Each table also includes the following information.

Class: Class [path] in IBM Administration Console for Content Platform Engine that you follow to access the property that is being indexed. If a class [path] is not listed, the property is a system property.
Database table name: Location of the property or properties that are being indexed
Database column name: Name of the property or properties, as they are displayed in the database
Content Cortex property name: Symbolic names of the Content Cortex properties that make up the index

Custom properties and the “uxy” prefix

Custom properties are easy to identify because they always have a prefix of either “uxy” or “uvxyz” attached to the beginning of the property name (where “ xy” and “vxyz” are automatically generated prefixes that help to identify a custom property).

Custom properties can be either user-defined or specific to an installed expansion product. Most databases name a custom property column by using the property symbolic name with a prefix. The prefix always begins with the letter “u”, which is followed by two or four hexadecimal characters, such as u3f_documenttitle or u3f1a_documenttitle.

Important: Custom properties that were created before the release of Content Cortex 4.5.1 have two hexadecimal characters that follow the letter “u”. Custom properties that are created after the release of Content Cortex 4.5.1 have four hexadecimal characters that follow the letter “u”. Content Cortex systems upgraded from earlier versions can have custom properties with both types of prefixes.

Tip: For more information, see Locating the in-use custom “uxy” property column.

Knowing the tools for creating Content Cortex indexes

You can use IBM Administration Console for Content Platform Engine to create single-property indexes or composite indexes using more than one property, unless the indexes are for a Content Cortex system property. If an index is for a Content Cortex system property, use the database management software to create the index. Use the Microsoft SQL Profiler, the Oracle Statspack, Oracle ADDM, or the DB2® Design Advisor to identify extra properties (columns) that you want to index.