After six years of gestation, it's finally here
Hadoop hits first 1.0 milestone
It's a big year for Big Data, and the specialists sent an early New Year's gift to us all over the holidays - releasing the first milestone of Hadoop.
The first major version of the MapReduce framework and computing platform is from the 0.20-security code line and adds security, webhdfs and HBase support (append/hsynch/hflush, and security) from December's 0.22.0 release. There is also the promise of enhanced access to local files for HBase plus the usual bug fixes and tweaks. All details can be found within the release notes.
Hadoop has now pretty much become the de facto standard for handling large data sets or complex processing, growing rapidly as some of the world's biggest companies backed the yellow elephant for their solutions. Some of the big names onboard and actively shaping Hadoop include Facebook, StumbleUpon, eBay, LinkedIn, Twitter, IBM and Adobe. No wonder it's become the hot topic within the Java ecosystem heading into the calendar year.
Redmonk analyst James Governor recently predicted big things for Big Data, especially Hadoop saying -
We’ll see a broad-based acceptance that Java is crucial to Big Data, and Big Data is going to be huge; that is, the Java renaissance starts here. Web companies built the technologies, but the enterprise is going to consume them. Hadoop is going to be stack-defining, and utterly pervasive.
There wasn't a bigger indication of Hadoop's dominance than Microsoft's decision in November to put the kibosh on their data crunching project Dryad, opting to join the open source revolution of Hadoop.
Its success since being contributed to the Apache Foundation in 2006 has spawned many companies devoted to the area such as Cloudera and Hortonworks and also further open source projects like Cassandra and Mahout, each developing their own communities as the word spreads. Now with a first major release, it seems that Hadoop will only go from strength to strength as the year goes on.