Jackrabbit Oak - Cutting diagnostic builds

Jackrabbit Oak as part of JCR implementation provides support for observing content changes via EventListener. Event listeners are notified asynchronously, and see events after they occur and the transaction is committed. EventListener can provide a filtering criteria for the type of events they are interested.

Event processing consume system resources hence its important that EventListener can tell Oak precisely which type of event changes they are interested in. The filtering criteria can be specified in following way

OakEventFilter

@since Oak 1.6

To make use of new filtering capability use following approach

+ +

import org.apache.jackrabbit.api.observation.JackrabbitEvent;
+import org.apache.jackrabbit.api.observation.JackrabbitEventFilter;
+import org.apache.jackrabbit.api.observation.JackrabbitObservationManager;
+import org.apache.jackrabbit.oak.jcr.observation.filter.FilterFactory;
+import org.apache.jackrabbit.oak.jcr.observation.filter.OakEventFilter;
+
+Session session = ...
+EventListener listener = ...
+
+ObservationManager om = session.getWorkspace().getObservationManager();
+
+//Cast to JackrabbitObservationManager
+JackrabbitObservationManager jrom = (JackrabbitObservationManager) om;
+
+//Construct a JackrabbitEventFilter
+JackrabbitEventFilter jrFilter = new JackrabbitEventFilter();
+
+//Wrap it as OakEventFilter
+OakEventFilter oakFilter = FilterFactory.wrap(jrFilter);
+
+oakFilter.withIncludeSubtreeOnRemove();
+//Set other filtering criteria
+
+jrom.addEventListener(listener, oakFilter);
+

Refer to OakEventFilter javadocs for more details on what filtering criteria are supported

Benefits of Filter approach

In Jackrabbit Oak JCR EventListener are implemented as Oak Observer which are backed by an in memory queue. Upon any save call done at JCR layer NodeStore in Oak pushes repository root node in this queue which is later used to compute diff from older root and then generate the JCR Events.

For events to work properly it needs to be ensure that this in memory queue does not get filled up. Having a precise filter helps Oak in prefiltering changes before they are added to queue. For e.g. if filter stats that its only interested in changes under ‘/content’ then Oak can check if changes under that path have happened for given change, if not then such a change is not queued.

Observation queue overflow

Why Oak

Jackrabbit 2.x is a solid and feature-rich content repository that works well especially for the needs of traditional web sites and integrated content management applications. However, the trends in user expectations (especially for personalized, interactive and collaborative content), application architectures (distributed, loosely coupled, multi-platform solutions with lots of data) and hardware design (horizontal rather than vertical scaling) have rendered some of the original Jackrabbit design decisions (which date back almost a decade) obsolete and there is no easy way to incrementally update the design.

Jackrabbit Oak aims to implement a scalable and performant hierarchical content repository for use as the foundation of modern world-class web sites and other demanding content applications. The repository should implement standards like JCR, WebDAV and CMIS, and be easily accessible from various platforms, especially from JavaScript clients running in modern browser environments. The implementation should provide more out-of-the-box functionality than typical NoSQL databases while achieving comparable levels of scalability and performance.

Work in Progress

This documentation is still work in progress. Currently much of the information on Oak is somewhat spread over different places. If you don’t find something here you could also try

- -

the Oak JIRA, specifically OAK-14, which lists the known backward compatibility issues,
the Oak development list,
the README files,
the Oak presentation from the .adaptTo conference 2012,
Oak, the architecture of Apache Jackrabbit 3.
Oak, the architecture of the new Repository

Jackrabbit Oak -

Oak Documentation

Observation

OakEventFilter

Benefits of Filter approach

Observation queue overflow

Why Oak

Work in Progress