Search Database Wizard

Because a single BRS file in the BRS (search) index cannot exceed two gigabytes in size, libraries with very large search databases should segment their databases. The Search Database policy is used to break data into smaller segments.

Segmented databases is a configuration option and should only be used for libraries with very large collections. Contact SirsiDynix Customer Support for information about configuring segmented databases.

The Search Database policy establishes a family name for the group of segments and specifies the number of segments created for this group. A segment list matching the specified number of segments is created internally at the time the Rebuild Text Databases report is run. The segmentation divisions are based on a publication year and catalog key combination. The Search Database policy determines the total number of segments to be used. Bibliographic records that do not include a publication year in the fixed field have a publication year value of zero. The Search Database policy sorts these records first. The list is then broken into the designated number of segments, minus one, based on the sorted publication year. The last segment is left free for expansion.

From the List Policies window, you can create, display, modify, copy, or remove the Search Database policies. Click Close to exit the wizard.

For example, for the UNION database, there are four UNI index segments. When the Rebuild Text Databases report is run, the data is automatically divided into four segments using a catalog key and publication year combination.

The Name attribute uniquely identifies the Search Database policy. This name is ten characters or less, and may not include spaces or punctuation, except for dash (-), underscore (_), and dollar sign ($). Additionally, the pipe character (|) may not be used.

This attribute provides more information about the policy and its use by the library. The description may be up to 60 characters in length. Although the Description attribute may contain spaces and punctuation, the pipe character (|) cannot be used.

The Index Name attribute is a 1-4 character field that uniquely identifies the index to be segmented.

The Number of Segments attribute allows you to type the number of segments this index should be divided into.