This documentation relates to an old version of DSpace, version 3.x. Looking for another version? See all documentation.
DSpace offers two options to index content for Browsing & Searching:
- Traditional Browse & Search (via Lucene & Database indexes) - this is enabled by default
- Faceted/Filtered Browse & Search (via Solr & DSpace Discovery) - disabled by default
This particular page only describes the "Traditional Browse & Search" indexing processes. For more information on Faceted/Filtered Browse & Search, please see DSpace Discovery, in particular .
Creating the Browse & Search Indexes
To create (or recreate) all the various browse/search indexes that you define in the Configuration Section there are a variety of options available to you. You can see these options below in the command table.
Arguments short and long forms):
Should we rebuild all the indexes, which removes old tables and creates new ones. For use with
Execute all the remove and create SQL against the database. For use with
Actually do the indexing. Mutually exclusive with
Write the remove and create SQL to the stdout. For use with
Create the tables only, do no attempt to index. Mutually exclusive with
Make the tables, and do the indexing. This forces
Print extra information to the stdout. If used in conjunction with
Delete all the indexes, but do not create new ones. For use with
Show this help documentation. Overrides all other arguments.
If you are using the Solr Browse DAOs it is not required to run this script as the data are stored in the Solr search core that need to be recreated using the Discovery maintenance script
Running the Indexing Programs
Complete Index Regeneration
Requires that you stop Tomcat first
Because this command actually deletes existing Browse Index tables, you must stop Tomcat (or your Servlet Container of choice) before executing
index-init. After the indexing command completes, you can restart Tomcat.
Known Oracle Issues
In many Oracle based DSpace installations, index-init often malfunctions because of Oracle specific permissions. It is therefore advised to stick to index-update instead
[dspace]/bin/dspace index-init you will completely regenerate your indexes, tearing down all existing tables and reconstructing with the new configuration.
Updating the Indexes
[dspace]/bin/dspace index-update you will reindex your full browse & search indexes without modifying the DSpace table structure. (This should be your default approach if indexing, for example, via a cron job periodically). Because it does not "tear down" the existing tables, this command can be run while DSpace (and Tomcat or similar) is still running.
If you are using the Solr Browse DAOs you don't need to run this script as the data are stored in the Solr search core. You need to recreate the indexes using the Discovery maintenance script
Destroy and Rebuild Browse Tables
This is really not recommended unless you know what you are doing.
You can destroy and rebuild the database, but do not do the indexing. Output the SQL to do this to the screen and a file, as well as executing it against the database, while being verbose.
At the CLI screen:
Browse Index Customization
DSpace provides robust browse indexing. It is possible to expand upon the default indexes delivered at the time of the installation. The System Administrator should review Browse Index Configuration to become familiar with the property keys and the definitions used therein before attempting heavy customizations.
Through customization is is possible to:
- Add new browse indexes besides the four that are delivered upon installation. Examples:
- Specific subject fields (Library of Congress Subject Headings). (It is possible to create a browse index based on a controlled vocabulary or thesaurus.)
- Other metadata schema fields
- Combine metadata fields into one browse
- Combine different metadata schemas in one browse
Examples of new browse indexes that are possible. (The system administrator is reminded to read the section on Browse Index Configuration )
- Add a Series Browse. You want to add a new browse using a previously unused metadata element.
webui.browse.index.6 = series:metadata:dc.relation.ispartofseries:text:single
- Note: the index # need to be adjusted to your browse stanza in the _dspace.cfg_ file. Also, you will need to update your Messages.properties file.
- Combine more than one metadata field into a browse.You may have other title fields used in your repository. You may only want one or two of them added, not all title fields. And/or you may want your series to file in there.
webui.browse.index.3 = title:metadata:dc.title,dc:title.uniform,dc:relation.ispartofseries:title:full
- Separate subject browse.You may want to have a separate subject browse limited to only one type of subject.
webui.browse.index.7 = lcsubject.metdata:dc.subject.lcsh.text:single
As one can see, the choices are limited only by your metadata schema, the metadata, and your imagination.
Because Browse Indexes are stored in database tables, remember to run
index-init after adding any new definitions in the
dspace.cfg to have the indexes created and the data indexed.
Since DSpace 3.0 a Solr DAOs implementation of the browse engine is provided. If you are using the Solr DAOs you don't need to run the script described in this page but instead use the Discovery maintenance script. Browse indexing in Solr is done within the Search Indexing process.
Search Index Customization
For information about configuring new Search Indexes, please refer to Configuring Lucene Search Indexes.