Blog Post

More Couchbase and Java with Views, Hadoop and Mobility

Raghavan Srinivas Published

I spoke at the London JUG and was probably over ambitious in trying to include our recent work on Views, Hadoop Connector and the Android Grocery Sync application in a single presentation besides demoing the "simple, fast and elastic" theme that Couchbase and the client libraries support. Each of these are exciting projects (and a presentation on its own right) from a Java developer perspective.

The aspect of being able to synch. data between mobile clients and servers via the Couchbase Mobile Synchpoint is certainly a sought after feature by mobile and Java developers.

The Hadoop Connector provides access to the Hadoop ecosystem and being able to use Couchbase as an operational store while tapping into the vast analytics capabilities in the back end enables a balance between meeting critical time-based needs using Couchbase and mining the data for intelligence via Hadoop.

Most importantly, the Views brings in a number of features including secondary indexing, filtering, ordering and even aggregating data. They are defined on the server for JSON-based data using Javascript-based Map-Reduce functions. The data is itself stored for efficient access using B-trees as outlined in the CouchDB: A Definitive Guide or in this article. From a Java developer perspective, it's very simple to access these views using the View, Query, ViewResponse and ViewRow objects as outlined in the Java tutorial which has been enhanced for views.

Overall, I am excited by the feedback I am getting from Java developers. There is still a lot of work remanining, but, I look forward to talking to more Java developers about NoSQL in general and Couchbase in particular.