Hadoop, the open source distributed computing platform, has some pretty impressive potential. But practically speaking, it's only as powerful as it is usable, which explains the recent introduction of a variety of products designed to simplify programming for Hadoop and related technologies. The newest, from JNBridge, appeared earlier in October. Here's what it does, and where it fits into the open source channel.
On Oct. 10, 2012, JNBridge announced the availability of a "lab" for connecting LINQ (Microsoft's Language Integrated Query framework) providers to HBase, a Big Data storage system developed as part of Apache's Hadoop project. Traditionally, because HBase is based on Java, interacting with it required the use of Java APIs.
Now, however, the interoperability toolkit released by JNBridge will make it possible to perform HBase queries using the LINQ interface from clients written in the .NET language. The lab can be downloaded from the JNBridge website, and requires the JNBridgePro product to work.
Hadoop: The Need for SimplicityUnless you spend your days writing code -- or talking to people who do -- the two preceding paragraphs may have read like Greek to you. But what this news means in somewhat less geeky form is that it will now be easier for programmers using JNBridge's software to write software for Hadoop without having to become experts in the particular languages associated with Hadoop itself. Instead, they can use programming languages they already know -- in this case the massively popular .NET toolkit from Microsoft.
And while that may not appear to matter much outside of the programming world, the niche that JNBridge and vendors like it are trying to fill could have much broader implications, especially in the open source channel.
In the era of the cloud and Big Data, Hadoop is rapidly emerging as a vitally important open source technology for building the next generation of software applications, and for addressing problems that have not yet even come fully into view. But Hadoop is also a massively complex platform, which means that learning its technical ins and outs requires a greater investment of time and intellectual energy than many IT staff are likely to be able to afford.
That's why solutions that make it easy to write programs that interact with Hadoop without having to learn advanced Hadoop code itself are likely to become a major focus going forward, particularly among open source vendors. The JNBridge lab is only the latest example of this trend. Pentaho is focused on the same niche, and Cloudera has also been releasing products for simplifying Hadoop deployments.
For other partners in the open source channel, there remain plenty of opportunities for addressing the usability hurdles that limit the ability of organizations to harness Hadoop's power fully. Expect this to become a crowded market as Hadoop and Big Data continue to evolve.