With your big data comes the need to secure it. How awful would it be if your competition got a hold of your data and used it against you, to their benefit? A lot. In Thor Olavsrud’s Nov. 12 column in CIO.com, he delves into some of the things that you, as a company using its data, need to know to keep it safe.
Many people start their projects or businesses, and only AFTER they’ve finished that, do they think about how to protect one of their most important assets. Thor’s analogy fits this way of thinking perfectly. He says, "You don't lock your doors after you've already been robbed, and you shouldn't wait for a data breach incident before you secure your data." Truth. Talk about it first with your team, then start building.
Figure out what data will, or even might, get stored. Make sure it is compliant. You don’t want to get busted later on down the line, lose your data AND money. If you do that, you’re gonna have a bad time.
Centralise accountability. Make your enforcement of policies consistent.
Encryption keys and encrypted data. To quote The Offspring, you gotta keep em separated. Would you keep the keys to your safe in the drawer next to the safe? That’s what I thought.
Use the Kerberos network authentication protocol. ‘Nuff said.
Secure automation. Since we’re dealing with multi-node environments, deployment consistency can be a bit tricky to ensure. You can use Chef and Puppet to stay up to date on things like patching, application configuration, updates, and more.
Add logging to your cluster.
Last but not least, of his suggestions anyway, secure communication between nodes and apps.
For more details, read the piece.