Global Warming hits OSS

“Scientists have been warning about global warming for decades. It’s too late to stop it now, but we can lessen its severity and impacts.”
David Suzuki.

We’ve all heard the stories of global warming leading to increasing occurrences of cataclysmic weather events.

Further to yesterday’s blog about big data and the ever-increasing data flowing through our B/OSS, have you noticed that the frequency of event storms also seems to be increasing? In fact, I’ve heard of some organisations who have so many events from their network devices, firewalls, intrusion detection/prevention mechanisms, etc that they simply can’t monitor all events. In fact, you could say that they’re in a state of permanent storm.

I know of one organisation that had so many events cycling through their monitoring consoles that they couldn’t even read any of them, let alone process them in any way. They asked their vendor to create filters that left them with a much smaller sub-set of events. These filters successfully suppressed (but didn’t resolve) a massive amount of low and medium risk events (at least they hope they’re lows and mediums). Unfortunately there are still too many high risk events for their team to handle. They got to the point of not even bothering to try to stem the flow, leaving the system running only for the purpose of ticking the “monitoring tools are operational” box on their risk evaluation audits.

When they increase the virtualisation in their network shortly, they’ll have a massive increase in devices under management. Fingers crossed that their “hope it doesn’t break” approach works out for them.

I wonder whether this OSS global warming scenario requires a whole new approach? Instead of processing on an event-by-event basis, can artificial intelligence / learning (AI) engines constantly monitor and refine their analysis based on trends and patterns in the events, perhaps even learning from event patterns across multiple customer networks? I know the revenue-assurance / fraud-detection sub-section of B/OSS has been using AI for years, but I haven’t come across it in advanced stages for event detection.

I’d love to hear from any readers who know of such AI products being available on the market already or coming to market soon.