DataNucleus database driver

The new DataNuclues database driver has improved many folds and supports industry leading RDBMS, HBase, NoSQL database and various document formats. What this means is that DataNucleus can be easily used with Hadoop Big Data applications and will also support pluggable variety of databases. Means that switching database is simple, by just changing the configuration for DataNucleus database driver.

DataNuclues supports JPA standard which makes it compatible with JEE as well. The driver is provided with Apache license, which makes it a good choice for variety of applications.

Tejas Bavishi

Google Map API

The Google Map API has advanced really well. The features include - Places Auto complete API, Distance Matrix API, Latitude API. This APIs also include 360 degree view of streets where available.

Places Auto complete API can be used in Location Based Services where user just types few initial characters and rest of the place name will be automatically filled. And the location of that business or restaurant will be automatically displayed on map.

The Distance Matrix API allows user to know distance between 2 places and what is the best way to go from Place A to Place B.

The Google Map Database is getting richer everyday and more correct and precise maps are released everyday.

For making location aware applications, developers have a brilliant opportunity to use Google Map API.

I am yet to see a comparison of different Map APIs like Google, Bing, NAVTEQ. But looking at Google Maps popularity and richness of APIs, Google seems to be ahead.

Tejas Bavishi

iPhone vs Samsung court room battle

My vote goes with Samsung, as there is nothing in terms of invention as to what Apple iPhone does or looks like. It is a sophisticated piece of technology but cannot be said that it is a invention. Just like microwave oven, refrigerators are not invention and so Samsung smart phones are not infringement to what Apple produces.

Tejas Bavishi

JEE 7 targets cloud computing technic

The new JEE 7 version is targetting cloud computing technic in a big way. The JEE 7 development kit includes cloud profile which will cover PaaS and some features of SaaS. What this means is that JEE7 will provide JEE compatible enterprise cloud APIs to develop web based Enterprise Java applications on cloud. These APIs also make compatibility to deploy on multiple platforms like Amazon, Eucalyptus and Oracle public / private clouds.

The PaaS feature includes features like WAR file compatibility on multiple cloud environments, scalable database connectivity, scale up and down of virtual machines.

Some of the SaaS features include multi-tenancy and database sharding. The JEE7 is due to release end of 2012.

Cloud infrastructure will be de-facto future standard for all enterprise applications. Just like internet banking is current way of doing things in financial industry. The benefits of a on-demand scalable cloud infrastructure weigh to much in favour of utilizing cloud than utilizing conventional infrastructure approach.

JEE7's cloud applications development features will take Java platform several steps in direction towards taking advantage of cloud computing. Also, adding support in JEE7 development kit will de-mystify SaaS cloud application development and vendor lock-in. It will also mean that IDEs like Eclipse and NetBeans will have support built-in for various cloud providers like Amazon EC2, Oracle, etc.

Java Architects and Java developers should keep an eye on new JEE7 release and also ISVs planning their roadmap for future product releases should cover their planning for JEE7.


Tejas Bavishi

Apache Hadoop as top batch processing framework / Solr indexing

Hadoop is increasingly gathering attention from Open Source Java community as top rated, highly scalable, in-expensive, batch processing framework. Some of the advantages of Hadoop are as follows:
(1) Apache Open Source
(2) Very high community support
(3) In-expensive
(4) Works on standard Windows / Unix server hardware, no special hardware requirements
(5) A company - Cloudera.com, providing training, commercial support and solutions.

Based on above advantages, Hadoop has become a very good tool for creating huge scale business applications that can process several peta bytes of data. In addition to this, Hadoop can also be integrated with database like HBase. More recently, Cloudera offers solution for connecting Hadoop jobs to Oracle database as well.

Hadoop integrated very well with Apache Solr for a complete indexing of documents for large scale search solutions. There are many references of integrating Solr with Hadoop on internet. Hadoop job can be used to upload data to Solr using SolrJ Java driver. Apache Solr search engine stores indexes on separate servers. A separate front end application can provide search interface directly from Apache Solr using SolrJ java driver to interface to Apache Solr. So on one side Hadoop uploads data (or documents) to Apache Solr, and on other side Java front end can provide search interface to Apache Solr.

Hadoop is definitely the technology of future and something all Java Architects should keep an eye on.

Tejas Bavishi