Predictive analytics has been around for a while and is applicable in several forms, most notably in business intelligence (BI) where you see it applied in areas such as Web traffic and consumer data analysis. It can also be applied to IT operations such as networking.
“Predictive analytics for IT is about understanding vast amounts of real-time data to forecast performance issues before they affect users,” said Daniel Heimlich, vice president at Netuitive.
“It is powered by behavior learning technology which is a math-based approach that involves advanced statistical analysis and algorithms.”
Netuitive, for example, has built a predictive analytics software platform based upon its nine behavior learning patents. It sits on top of the enterprise IT infrastructure stack where it collects and analyzes data in real-time from monitoring tools from vendors such as BMC, IBM, NetApp, CA, HP, VMware, Microsoft, Oracle and Compuware. Heimlich said this enables end-to-end monitoring and management of complex infrastructures with a very small footprint in terms of storage and computing resources.
Virtualization Heightens Analytic Needs
A greater analytic approach is needed these days, particularly in light of the rise of virtualization. This has changed the conventional IT systems management game. The inherent dynamics of virtualization made traditional base-lining and rules-based approaches to systems management far more challenging, if not obsolete, especially in large heterogeneous environments.
“Virtualization and cloud environments allow little visibility into resource utilization and root cause analysis, causing great concern for application owners,” said Heimlich. “While it has been around for years, it was not until the advent of virtualization that it found its home.”
Now virtualization is moving up to the next level. Cloud computing has seized upon virtualization to propel IT towards the vision of IT as a Service (ITaaS). As organizations virtual more and more resources, they can move towards the virtualization of all the services IT provides based upon a cloud model. And since virtualization underpins most cloud solutions, the role and benefits of predictive analytics as the basis for management in the cloud are the same.
“But as the nature of the virtualized environment becomes even more dynamic, complex and real-time in the cloud, the benefits of IT analytics become even more dramatic,” said Heimlich.
He commented that this coalescence of trends is showing up sharply in the marketplace. Netuitive has seen a rush of large enterprise customers who want to manage complexity that they say is now several orders of magnitude higher than it was just a few years ago.
“They are starting to achieve visibility across platforms and vendors using a math-based approach involving behavior learning and predictive analytics,” said Heimlich. “These IT analytic-based approaches are being validated by big early adopters in financial services and telecommunications who are some of the largest and most demanding deployments of virtualization in the world.”
He expects 2011 to be a breakout year for management in the cloud fueled by this framework of analytic-based approaches. IT leaders, he believes, will start to realize the promise of the cloud by serving line of business owners more efficiently, CFOs will realize lower hardware costs and energy bills from right-sized infrastructures, and application owners will be able to deploy or change resources in minutes, not weeks.
“This is what effective virtualization and cloud management is all about — service-level visibility, automated problem diagnostics and predictive analytics enabling organizations to manage their performance and capacity proactively and end-to-end,” said Heimlich.
He cites some impressive numbers to back up his claims. The Netuitive user base includes seven of the top 10 banks and two telco giants who are now able to predict degradations and avoid outages for their most critical applications.
In one case, the organization is using predictive analytics to analyze more than a million metrics simultaneously allowing it to eliminate 3,480 hours annually in service degradation representing a business savings of $18 million.
David Williams, an analyst at Gartner studied this implementation. He said that this firm, one of the world’s largest telecommunications companies, became an early adopter of behavior learning management software, which, by leveraging and enhancing existing management tools, moved from a reactive state to a proactive one where IT issues are identified before impacting service levels.
“A large telecommunications company achieved significant, measurable cost savings and increased IT operations management efficiencies through the adoption of a new, emerging availability and performance management (APM) solution using behavior learning technology,” said Williams. “Behavior learning technologies have matured, allowing IT organizations to move from a reactive state focused on mean-time to repair to a proactive state where outage avoidance becomes a more realistic objective.”