3 Factors to Consider When Securing Big Data

Big data is the new toy in town—a technological commodity that is driving development, but is also a major point of contention between companies, users, and governing entities. But despite the name big data, it is often in the possession of small businesses, who have not taken the appropriate measures to secure this data.  When such large amounts of information are on the line, a breach of this data can be extremely detrimental.

With continual scandals being aired concerning poor privacy protections, it is even more important for your data to be protected. Consider these three things when securing big data: your specific configurations, what access you give out, and how to monitor your data.

1.  Configurations 

It was June of last year that the Exactis leak was revealed. Exactis, a Floridian marketing data broker, had a misconfigured Amazon ElasticSearch server that exposed close to 340 million records on both American adults and businesses. This included incredibly specific details such as pets, gender of children, and smoking habits. This leak has crippled Exactis; there is little chance that Exactis will bounce back from this event.  Beyond the effect that this leak has had on the business, Exactis CEO, Steve Hardigree, has also been open about the stream of inquiries, threats, and constant stress this has had on his personal life.

The root of this crippling leak lies in a misconfiguration and shows us just how configurations can make or break your business.  When you are planning out your big data space, you need to double, and triple check your configurations.

Tips for Checking your Configurations:

  • Security is a multi-layered beast and your data is unique, which in turn means that your approach to security must be customized. This could mean using security software in an unconventional manner or utilizing a third-party security company.
  • Think of the little things. Do you trust all of the programming interacting with your data? If not, how can you make it a trusted resource?
  • Consider getting a third-party Network Security & Architecture Review of your environment. This allows you to have an outside opinion of exactly how secure your data is. If possible, it is beneficial to get this review at least annually.

2.  Access Granted

As you are deciding on configurations, you need to take into account who will be granted access and to what.

If the data is meant to stay completely internal, you need to decide what kinds of users are allowed what permissions. For example, who is allowed to pull data? Is anyone? If it’s not a part of the daily workload, under what circumstances is it allowed? By who?

If you are going to share your data with third parties, there is another host of questions to consider.  Do you allow them unlimited access to your data? Who do you allow access to?

Tips for Granting Internal & External Access:

  • Limit the amount of external access you allow; if possible, do not allow it at all. This will lessen your attack surface and your inherent risk.
  • External resources likely don’t need to access everything your internal resources can. Restrictive groups are a great organizational way to separate who has access to what within your environment.
  • Not all internal resources are equal and therefore should not be given the same access. You will need to evaluate how you give out access and document your process of escalating and deescalating access.

As it has become evident with Facebook’s admittance of leaving data connections open even after deals had been closed, it is also important to think about what happens when access has been revoked. What are you going to put in place to prevent access when it should no longer be allowed?

Take the access you grant seriously so you don’t end up scrambling to make changes after an incident.

3.  Monitoring & Alerting

For everything that can be done to your data, there should be a way for you to monitor it. That is not to say that you have to micro-manage every aspect of your big data. But if an incident were to occur, or more realistically when an incident occurs, you should be able to construct an image of what was going on at the time of the event. For this to be possible, you need a way to monitor your data and receive alerts on the incidents.

Tips for Monitoring & Alerting:

  • Adversaries do not keep normal business hours, so be sure you are monitoring your data at all hours. One way to easily achieve 24/7/365 monitoring is by outsourcing this function to a Managed Security Services Provider (MSSP).
  • When setting up alerts, it can be challenging to find a balance between “alert on every single possible event” and “I only want to see important alerts”. What if an uptick on those seemingly harmless alerts is the only tip-off to an insider threat? And on the other hand, if you are constantly on edge from alerts, you will easily fall into alert fatigue. An MSSP can act as the filter between you and your alerts, only notifying you after an alert is investigated and confirmed to be legitimate.

When you are in possession of big data, there is a lot on the line to secure.  When a breach of this magnitude can destroy your business, it’s critical you take into consideration these factors.