Regenerate Data Indexes

Each of the Regenerate*Index configuration parameters rebuilds one or more of the IDOL data indexes for optimized fields (MatchType fields, NumericType fields, and so on) from the textual document field values held in the server. You can do this to reconcile the index or indexes with the document data if:

In these cases, the alternative to regenerating is to reindex all documents. Regenerating is generally significantly faster, as only the requested indexes regenerate.

In addition, you might need to regenerate fields if you have indexed ReferenceMemoryMappedType fields out of order, with some child fields indexed before their parents. You regenerate these fields either by using the DREREGENERATE index action, or by using the Regenerate feature on the Service Control tab of the Console page in the Control section of IDOL Admin. For more information on the DREREGENERATE action, refer to the IDOL Server Reference.

You can regenerate indexes for the following field types:

Additionally, you can regenerate the following indexes if validation fails:

Regenerate an Index

To regenerate a data index

  1. Stop IDOL Server.

  2. Make any required changes to the [FieldProcessing] or [Properties] configuration sections.

  3. Set the required Regenerate*Index settings to True in the [Server] section of the configuration file.

  4. Restart IDOL Server.

At start-up, IDOL runs through all stored documents, and loads the values of any fields to regenerate from disk. It inserts these values into new index structures.

When the process is complete, IDOL Server deletes any existing indexes and replaces them with the newly built versions.

After the server has started, remember to remove or turn off any Regenerate*Index settings in the configuration file if you do not want the server to regenerate the indexes every time it starts.


Regenerating the index requires a server restart. In addition, regenerating indexes also results in an increased startup time.


Regenerating the unstemmed index is an intensive process and generally takes significantly longer than regenerating other indexes (though it is still faster than reindexing the data).


Consider the following points when you are using the regeneration options: