Neo4j 1.2: What’s New

Neo4j 1.2 was released on December 30th. Now that’s a very weird time to make a major release. But according to the Neo4j roadmap and milestone reports, Neo4j 1.2 brings quite a few major changes and improvements.

First major shift in Neo4j direction is that it is now available as a RESTful server. Even if a Neo4j REST API existed before, this shift from promoting an embedded graph database to a full blown RESTful graph database was firstly announced with the first 1.2 milestone. As someone suggesting this change, I cheer the decision.

The second major feature is the high availability Neo4j cluster. Most of the existing graph databases have started their life as embedded storage solutions. Then a few of them have seen the light of becoming server-based storage solutions. But with that also came questions related to availability and scalability.

Starting with this version, Neo4j offers the option of setting up a high availability cluster and this is a major step forward for graph databases. This is still a first version where writes are slower, the cluster is not elastic, and there are limitations at the distributed transaction layer.

Scaling graph databases remains a very complicated problem to be solved. Darren Wood’s¹ presentation covers some of the challenges of distributed graph databases

Neo4j 1.2 features a couple of more goodies like a smaller footprint kernel and an automatic JMX enabled monitoring and management component.

The original announcement covers more details about this major Neo4j new version. The only missing piece from this release and announcement is a document describing Neo4j API changes. But that should not stop you from trying it out.

Original title and link: Neo4j 1.2: What’s New (NoSQL databases © myNoSQL)