I love Hadoop for what it does. One advantage to it: it enables the repositioning of common data manipulation tasks to distributed points on a network, opening up the ability to perform analysis across multiple data sources across various web-enabled sources and leveraging multiple data resources as though they were integrated into a single autonomous data store. That’s one of its advantages.
But Hadoop is still relatively new. So it amazes me that end-users today are shocked, shocked I tell you that infrastructure capabilities such as robust security and multi-user access haven’t been fully implemented.
Information Week magazine published a great piece in which Robert Bird of Red Lambda commented that Hadoop “really isn’t designed to be a secure processing environment, which is a little scary considering how many people are trying to use it” that way, adding “[w]e see Hadoop being used to solve one problem here and two problems there … [w]hat we don’t see is 75 or 100 people in the environment all writing different programs and using it for this big cluster. We don’t see it providing the real economies of scale that it should at the data-center level.” [1]
It took Oracle and other vendors years to get those capabilities to the point of maturity. I think a lot of modern users expect to snap two fingers and viola, instant mature complex data analysis tool. Open source and crowd-based collaboration are certainly helping to speed the development cycle. But complex systems like Hadoop still need time to develop, at least – in the current environment.
Be patient, Hadoop-ians. It’ll happen.
Footnotes
[1] Hadoop Security: Some Enterprises Miss Risks. Information Week, August 8, 2012. http://www.informationweek.com/big-data/news/software-platforms/240005132/hadoop-security-some-enterprises-miss-risks1