Index areas and indexes

An index area is a file system directory that contains index files. One or more index areas can exist for an object store.

Each index contains indexing information for the objects that belong to the same indexable base class or to subclasses of the base class. The text in the content and CBR-enabled properties of an object are stored as one entry in an index. Consequently, a full-text search returns one search result for each object regardless of the number of times that the object text matches the search criteria.

Both an index area and an index have a status property that controls whether index information can be changed or added. (The status does not control the accessibility of index information for IBM® Content Search Services search servers. Index information is always available for searches.) The possible statuses are OPEN, FULL, CLOSED, and, for index areas, STANDBY.

Object indexing

An object might be indexed for the first time in the following circumstances:

When indexing an object, Content Engine randomly selects an OPEN or FULL index area. If a FULL index area is selected, and the object cannot be indexed in an existing index, another index area is randomly selected. If no OPEN index area exists, and no FULL index area exists with suitable existing indexes, a STANDBY index area is randomly selected. The status of this index area is automatically changed from STANDBY to OPEN.

Whether an object can be indexed in an existing index is partly based on the status of the indexes and index partitioning. With respect to index status, a previously unindexed object can only be indexed in an OPEN index. A new OPEN index is created in the selected index area for the object if the index area has no suitable existing OPEN index.

Object reindexing

A currently indexed object might be reindexed in the following circumstances:

When an index or class is reindexed, IBM Content Search Services performs the following actions:

When a currently indexed object is reindexed, Content Engine finds the home index: the index that currently contains the index information for the object. One of the following actions is performed:

Index creation

The following list is a summary of the events that might trigger the creation of a new index:

Related concepts
For information about index partitions, see Index partitions.

Related tasks
For information about changing the status of an index area, see Changing the status of an index area. For more information about object store migration, see Migrating existing indexes into IBM Content Search Services.