Select Page

Can you predict the death of your network before it happens?

By Eli Sabatier

If I’ve said it before, I’ve said it a million times: data now rules the world. And trust me when I say I’m inundated with it on a minute-by-minute basis. From incoming orders, to outgoing shipments, to customer care—and don’t even get me started on our marketing team hitting me with dashboards 24/7. And with all that said, I wouldn’t have it any other way.

In fact, utilizing business intelligence (BI) data as a primary business driver is nothing short of revolutionary. To be able to see in real time how a business is running in any department with KPIs, and more, makes everyone’s life easier, even if we have to change habits to manage the new world view.

And with that type of insightful BI comes an entirely new way to peer into the inner workings of IT, but this time using predictive analytics to troubleshoot one’s own network. However, the first question should always be how well does it work?

For years, analytics tools in IT have been a bit of a slippery slope conversation. Yes, there have been some analytics that have worked very well: the numbers creating a baseline for predicting when and how IT infrastructure would need attention. A simple example is that of capacity. Calculating the amount of data being created by certain systems or inputs is an easy formula that shows how much capacity will be required by a projected date. But that is no different from calculating the amount of water dripping into a bucket, figuring out when another bucket needs to be added.

However, when it comes to predicting the lifespan and potential issues and failures with physical hardware, the math and the tools are very different. In this particular case, the analogy of the bucket changes from when another bucket must be added, to knowing when the bucket will develop a slow leak. That’s not math so much as it is predicting the future through a crystal ball, at least up until now.

Call it an outcome of digital transformation, the information age, big data, or any other industry buzzword you like. The reality is that predictive analytics can now calculate a multitude of network issues. This is made possible through statistical analysis and data mining, enabling tools to be “trained” through pattern recognition. And though I don’t claim to be a mathematician, the simple explanation is that the analytics tool calculates patterns that show regularities in the collected data, and then, subsequently, defines potential irregularities: the problems that could arise that could adversely affect the network.

So, by now I’m sure there are at least some of you (including myself) thinking back to the movie, Minority Report—a movie based on the idea that analytics can show someone who will commit a crime before they actually do. And though I hope that a dystopian society is a long way off, using analytics to keep my business running is my kind of science fiction.

In the case of network predictive analytics, when well designed tools are implemented businesses of all types can almost immediately reap the benefits. From identifying the aforementioned anomalies and potential hardware risks within a network, to also helping with more complex administration, the possibilities are endless.

For instance, in industries where lives depend on networks running perfectly at all times—transportation, healthcare, and so on—having a watchful eye that can predict outages before they happen is nothing short of incredible. It’s this type of predictive data that can ensure hospitals see fewer issues, air travel remains safer, and more.

Then, of course, there is the human element and how it can impact the safety of others. As I’ve written in many previous articles, we live in dangerous times as it pertains to cyber threats and crime. Having a tool that uses predictive modeling and analytics to search for anomalies including access, policy violations, or any other activity that is unusual, is certainly a giant step in the right direction.

And though this technology is nowhere near the point of helping me win my office bets on the winner of the next world series, this technology is emerging as something that will be commonplace in IT infrastructure in the coming years. My advice is to embrace the new world order of analytics on all fronts to make your business infrastructure better, more reliable, and safer. And, maybe, one day a world series prediction will come true. 😉