The race for dominance in the Hadoop security market niche is definitively on. Vendors of popular distributions and of security add-ons have started to recognize that clients long for a harmonized approach to Hadoop security that will make controls that are at first specific to the resource effective across a majority of Hadoop components. Two possible routes to this are the extension of the Apache Sentry or the adoption of a XACML-style authorization language. The market currently also argues about the right demarcation point for the security of your data in Hadoop: Will the Hadoop distribution eventually take care of the vast majority of security requirements, or will this become the domain of add-ons similar to the market for database audit and protection tools? And if the Hadoop security market niche is not absorbed by the bigger distributions, will the add-ons eventually also provide security for plain vanilla Hadoop that is part of some Linux distributions but has none of the features of the established Hadoop distributions?
My research note “Protecting Big Data in Hadoop” helps you navigating the lands of big data security and gives you guidance on how to use your choices wisely and what controls and add-ons to put on your pick list.
P.S. Access requires Gartner GTP subscription.