Do More With Less: Predictive Analytics For I&O
Moore’s Law was bound to catch up with us. Loosely applied, it says that technology grows more complex every year. Human brains do not. People can’t keep up with monitoring, debugging, and managing today’s technology. Users’ rising expectations make it even worse: they want features and fixes in minutes, not days or weeks. Technology may soon get away from us.
The American comic strip character Pogo put it this way: “we have met the enemy and he is us.” In this case, our enemy is also our best ally. Surely we can harness technology’s power to help us keep it under control. We can, we are, and we will. Predictive analytics, common for decades in other industries, is now a growing force for monitoring and managing business technology, and has the potential to put us back in control of our runaway technology.
The least sophisticated analytics predicts what instrumentation is appropriate for a server based on what software it’s running or what kinds of network traffic is going in or out. For example, is database software found, or are SQL queries going in and out? This analytics drives automation that reduces manual administrative work.
Moderately sophisticated analytics predicts trouble based on simple trends like CPU utilization rising, memory consumption rising, or free storage declining; and drives capacity planning before a resource crisis occurs.
Really sophisticated analytics watches multi-variate trends such as cycles of high user demand (for example monthly sales campaigns) coupled with performance expectations and resource constraints, to drive automated resource scale-up (to sustain best performance) or scale-down (to reduce over-provisioning costs).
In my thirty-plus years of technology management, I’ve never seen an advance this exciting, or with this much potential to enable I&O teams to do more with less. Predictive analytics doesn’t just put us back in control of our technology, it decisively wins the battle to maintain it – so we can shift our attention to more important things like customer obsession.
If you’d like to learn more, check out my recent reports on predictive analytics: BT Analytics Helps You Win In The Age Of The Customer and Predictive Analytics Can Liberate I&O Pros From The Tyranny Of Firefighting.
As always, your comments are welcome!