With Harvard Business Review calling the Data Scientist the sexiest career for the next 10 years, security professionals are going to have their hands absolutely full with securing the distributed infrastructure that powers big data analytics. The Hadoop infrastructure isn’t just one tool that you download to get you some Big Data fun, it’s really a framework of a multitude of tools (and options for substitution) that each carry out specific tasks in a distributed and flexible way.

Servers Stock Photo, by getButterfly

Part of the driving force behind wide-scale Hadoop environments is the notion that it is easier to move computation capabilities than it is to move data. This means that nodes will have some slice of data, but the end result analytics would really take into account all of the nodes and their respective slices. This opens up a ton of potential attack points whereby nodes of malicious intent can do bad things. It’s not just about stealing information or analytics, it could be spreading disinformation as well.

For example, imagine a competitor using its new big data cluster to help define product strategy. If end users blindly trust the data, minor manipulations by a malicious node could dramatically alter the outcome of the analytics, thus causing a company to place a bet on a guaranteed loser. Bad for that company, good for a competitor.

Security professionals are behind the curve of big data. First, we hate the buzzword and its misuse, and secondly, for most of us the clusters are not in wide-scale deployment in areas where we can influence. If you need a crash course on big data security, go check out Securosis’s whitepaper, Securing Big Data: Security Recommendations for Hadoop and NoSQL Environments. This will set up a baseline such that you can ensure you company will be able to productively use big data in big business decisions.

This post originally appeared on BrandenWilliams.com.

Possibly Related Posts:


Tagged with:

0