Remote Repository Access

The content repository within this web application is made available to remote clients through RMI and the jackrabbit-jcr-rmi component.

The remote repository stub is available both in the RMI registry (one is started automatically by this web application if not already running) and as a direct HTTP download. The default URLs for accessing the remote repository are:

Note that the above URLs are the defaults. You can disable or change them by modifying the /WEB-INF/web.xml deployment descriptor.

Accessing the remote repository

To access the remote content repository you need to use the jackrabbit-jcr-rmi component in your application. If you use Maven 2, you can declare the JCR and jackrabbit-jcr-rmi dependencies like this:

<dependency>
  <groupId>javax.jcr</groupId>
  <artifactId>jcr</artifactId>
  <version>1.0</version>
</dependency>
<dependency>
  <groupId>org.apache.jackrabbit</groupId>
  <artifactId>jackrabbit-jcr-rmi</artifactId>
  <version>1.4</version>
</dependency>

With that dependency in place, you can use either the RMI registry or the direct HTTP download to access the repository.

The required code for accessing the repository using the RMI registry is:

import javax.jcr.Repository;
import org.apache.jackrabbit.rmi.repository.RMIRemoteRepository;

Repository repository =
    new RMIRemoteRepository("//localhost/jackrabbit.repository");

The required code for accessing the repository using the RMI registry is:

import javax.jcr.Repository;
import org.apache.jackrabbit.rmi.repository.URLRemoteRepository;

Repository repository =
    new URLRemoteRepository("http://mastercard-uat-graph.spectrum.precisely.com:-1/RepositoryService/rmi");

See the JCR specification and the Repository javadoc for details on what to do with the acquired Repository instance.

Remote access performance

Note that the design goal of the current jackrabbit-jcr-rmi component is correct and complete functionality instead of performance, so you should not rely on remote access for performance-critical applications.

You may want to look at the Jackrabbit clustering feature for best performance for concurrently accessing the repository on multiple separate servers.