[resolved] Indexing in Talend MDM

[Anonymous]

Hi.
As I know eXist db supports several type of indexes.
How does Talend MDM manage indexes?
Is it possible in Data Modeler add indexes manually?
Thank you.

[Anonymous · Accepted solution]

Hi,
The data modeler does not support indexing (for the moment). It has to be done in eXist directly. You can check the following URLs :
- General Indexing information : http://exist.sourceforge.net/indexing.html
- Lucene (full-text) indexing : http://exist.sourceforge.net/lucene.html
By default, Lucene is activated for Talend MDM and the queries done in Talend MDM use this index.
Indexes are declared through a document in the database called "collection.xconf". To index a specific collection, put the document into /db/system/config/<same path as the collection>. For instance, I have a ?Location? data-container, it is stored in /db/Location. So I create the collection.xconf doc into /db/system/config/db/Location. Use the eXist DB client to connect to your database:
- Go into /db/system/config
- Create a "db" collection using File > Create Collection
- Go into /db/system/config/db
- Create a <collection> collection, where <collection> = same name as the collection to index in /db (for instance: ?Location?)
- Create a collection.xconf document in /db/system/config/db/<collection> using File > Create Empty Document
- Open collection.xconf and create the indexes
You can use both Full text and "n-gram" indexes (with "qname").
- Reindex the collection by selecting /db/<collection> and then: File > Reindex Collection
Regards.