Fork me on GitHub

Repositories Metadata Content Storage

With version 2.0.2, Metadata repository content can also be stored in an Apache Cassandra database.

It's possible to switch implementation with configuring the system property ( with one of the id below.

There are now 3 implementations of storage:

  • File (id: file)
  • Jackarabbit (default one) (id: jcr)
  • Cassandra (id: cassandra)


Prior to version 1.4-M1, repository content is now stored in a jcr repository (based on Apache Jackrabbit implementation).

A default Jackrabbit configuration is provided :

Error during retrieving content skip as ignoreDownloadError activated.

You can use your own configuration by adding a file repository.xml in ${appserver.base}/conf.

By default, the Jcr repository is stored ${appserver.base}/data/jcr.

If you want to change this default location, you must edit the file WEB-INF/applicationContext.xml, uncomment/edit lines and change with your values:

<bean id="jcr-config" class="org.apache.archiva.metadata.repository.jcr.ArchivaJcrRepositoryConfig" factory-method="create">
  <constructor-arg value="${appserver.base}/conf/repository.xml"/>
  <constructor-arg value="${appserver.base}/data/jcr"/>


With the distribution including the embeded Jetty, you can configure Cassandra runtime in the file conf/

The default content:

If you deploy in a Web Application container, you can configure a path to this configuration file using the system property: -Darchiva.cassandra.configuration.file=