Home / Educational Content / Database & Technology / SELECT Journal / Big Data Defense Is a 5-Person Game

Big Data Defense Is a 5-Person Game

By Ian Abramson

In sports, we talk about defense being a team game. One person alone cannot save a team from defeat; the entire team must work together to support a strong defense in order to win. It takes more than a hot goalie to win. The same can be said in the age of big data. One person and one method will not suffice to meet the security threats we are faced with today. The desire of competing companies who want access to information grows, because information is power, and the power struggle for information is unending.

How can we prepare our organizations for the information attacks and hacks we know are coming? We need to consider this more than ever as data architectures change and evolve. We no longer look to build the behemoth data warehouse but are instead deploying data lakes that can help support data scientists. Analysts look deeply into questions that had been difficult to ask when data was scattered around the organization. Today, we try to bring all of our data together in the data lake where we have created a repository that has more information than ever before. Thus, we are creating a new risk to the business. To address this risk, we need a multifaceted approach where we look internally and externally to protect ourselves and our data.

Internally, we must create an environment where data is respected and governed in ways which limit information access but do not limit the analyses that can be performed. We use methods such as predefined access methods that resemble APIs, as well as user access controls. In addition, we must establish a program of data governance to support these requirements. Governance will enable the business to understand what is available and how data can be used. It does not have to be an overbearing process; rather, it should be one that meets the basic needs of information sharing.

Externally, we must look at our big data security as deeply as we do with our mission-critical operational databases, securing the data lakes with multiple layers of controls to limit access to those who have the right authorization. Data should be encrypted at rest and in transit to ensure that even if someone reaches your data, it will be protected. Security in the Hadoop space is maturing, but we need to take the basic steps to make sure we keep out the bad guys.

With a coordinated approach to security and information defense, it will be possible to secure big data — your new most valuable assets — in a proactive way while enabling business analysis never dreamed of before.

Big Data Defense Is a 5-Person Game