By Deepak Advani
Imagine receiving one million emails per day, with no way to tell which ones actually need to be read and which ones can be ignored.
That’s essentially what’s happening with IT operational data in organizations around the world. Each day the IT network of a typical organization generates more than 1.3 terabytes of operations, including data such as log files, software error alerts, IT service tickets and network updates.
Taking a passive and reactive approach to these events is no longer acceptable. With an increasing amount of data directly related to how quickly we can adopt and scale mobile and cloud computing, greater knowledge of what this operational data means – and what it can predict -is required to not only keep the lights on, but to optimize performance.
The answer lies in analytics and cognitive computing. Over the past few years, the technology industry has made enormous strides in how we translate and learn from Big Data. We’ve applied cognitive intelligence to analytics methods to not only interpret data, but to clearly see and learn from patterns in our business operations, customer trends and physical infrastructures – just to name a few. And the results have transformed our industries, marketing and cities. So why not apply this same intelligence to our IT systems?
As valuable as it is, organizations have previously had no choice but to discard or archive their IT operational data. Many have missed the boat on key information – information that might have helped prevent system downtime, or helped them maintain power in their networks when they needed it most – such as during a major app deployment or a multi-billion dollar financial transaction. But enterprises have had no way to capture these insights, aside from hiring a team of data scientists. Even if they did so, it would be a poor allocation of resources.
As Glenn O’Donnell and Jean-Pierre Garbani, infrastructure and operations analysts at Forrester Research, put it in a recent report:
“The value of IT is not to just process data and perform repetitive ‘intellectual grunt work.’ All of us in IT need to turn our intelligence and creativity toward making our business more successful….let the robots do the hard work. Let humans do what humans do best: Create!” (Turn Big Data Inward With IT Analytics, Forrester Research, Inc.)
Starting now, we can fill this void by tapping into an entirely new category of analytics software: IT Operations Analytics. Combining analytics with cognitive computing, we now have the ability to sift through terabytes of operations data in real time, spotting and learning from trends critical to IT health and performance. New software, such as IBM SmartCloud Analytics – Predictive Insights, uses its cognitive capabilities to learn, reason and sense an organization’s IT systems. So as business and performance conditions change, our systems can adapt; updating settings and eliminating the inadvertent but costly errors of poor system configuration.
Let’s take an industry look. The Network Operations Center at Consolidated Communications, a Midwest U.S. telecommunications provider, monitors 80,000 network utilization metrics on their Internet, IPTV, and private networks. This information is derived from operational data from their network elements, but because of its volume, the company could only identify large disruptions in service.
Since implementing operations analytics, the company now uses cognitive intelligence to learn and predict normal usage and performance for each individual customer, automatically setting and monitoring the most important metrics. This means Consolidated Communications can define good service for every single client, and can immediately determine when critical deviations happen – no matter how small. Not only is the company able to save on reactive costs, but customers now enjoy extremely tailored and personalized service.
The need for analytics in our IT environments is urgent. Combined with natural learning abilities, analytics can take a complex IT environment overflowing with data and transform it, and provide us with the right information at the right time. By putting advanced analytics and predictive, cognitive tools to work in operational systems, we can now turn operational data into a competitive tool to not only incorporate mobile and cloud expansions, but to continue to evolve and manage our IT systems as our world’s innovation grows.