he National Security Agency (NSA) has been taking the heat for several months now, no doubt. And there has been plenty of commentary on the intelligence agency already. Although this blog entry is going to delve into the NSA, I’m not going to share my thoughts on the fundamental controversy of whether the NSA has overstepped its bounds. Instead, I am going to share some interesting details, on what the NSA has been doing with OpenStack. Frankly, it’s pretty neat stuff.

At the OpenStack Summit in Portland last April, the NSA pulled back the curtain on its private cloud. Nathanael Burton, computer scientist with the NSA, started the talk with what he could not share: the number of users, systems or servers on OpenStack, how much storage capacity the NSA had, which apps it uses, and his favorite color – all of which elicited hearty laughs and applause from the audience.

But he did share this. In Fort Meade, Md., where the NSA has its headquarters, IT experts like Burton work to protect and secure the federal government’s computing systems. There’s a mix of commercial, open source and in-house developed technology.

For some time, the NSA had been eyeing cloud technology. But like so many, there was skepticism. “In the media, cloud’s been this buzzword that’s been blown out to be everything, right?” Burton told OpenStack Summit attendees. “It solves everything. It’s the panacea.”

But Burton and his team knew that wasn’t the case, and said the NSA has its own cloud definition. “We think cloud is big data,” he said. “Obviously, we probably have a lot of problems that have a big data focus.” He said that for NSA big data is the ability take data, enrich it across other data sets and analyze it in ways that go beyond more traditional processes.

To support its big data projects, the NSA has developed Apache Accumulo, which is built on Apache Hadoop. In 2011, the NSA contributed the system to the Apache Foundation as an incubator project. As for cloud, Burton said, there were lots of stovepipe teams working on cloud, but there was no cross-component orchestration that would enable efficient service delivery.

Much of the NSA’s IT development work typically got mired in the approval process, which could take weeks or months. Burton realized something needed to change, and starting working on a plan. “The problem I was trying to solve was it took too much time from idea to capability to actually deploy anything in our environment, and we needed scale and we needed agility to be able to do things quickly, as missions changed and demands came up.” What Burton was looking for was lower barriers to entry, self-service, on demand, elasticity, and API access.

Source and Full story:
Going Big with OpenStack at the National Security Agency