Secondary Access Methods

The secondary access method handles all indexing operations for a particular data type. If the operator class of a secondary access method does not handle a particular data type, you cannot build an index on that data type.

Dynamic Server provides two built-in secondary access methods:

• Generic B-trees
• R-trees

Generic B-Tree Indexes

A B-tree index organizes index information. A B-tree index is arranged as a hierarchy of pages. Dynamic Server uses a B-tree index for the following values:

• Columns that contain built-in data types (known as a traditional B-tree index)

  Built-in data types include CHARACTER, DATETIME, INTEGER, FLOAT, and so forth.

• One-dimensional user-defined data types (known as a generic B-tree index)
• Values that a user-defined function returns (known as a functional index)

For more information on B-trees and functional indexes, see your IBM Informix Performance Guide.

R-Tree Indexes

The R-tree indexing structure supports spatial data. An R-tree index uses a bounding box, which is a set of coordinates that contains one or more objects and supports spatial data (two-dimensional, three-dimensional, and so on). An object can theoretically belong to more than one bounding box. An R-tree index is useful for searches on multidimensional data.

For information about R-trees, see the IBM Informix R-Tree Index User's Guide.