How collective intelligence is reshaping systems management

Big data has always had a place in the world of systems management, but it might have found its sweet spot in the cloud. While there are plenty of tools available for analyzing data on how your physical resources are operating in your data center, it’s still a lot of work and it’s not easy to truly figure out what’s going on. You could think of it like a Yakov Smirnoff joke: In data center, you discover insights. In cloud, insights discover you.

What that means is that companies providing cloud computing and cloud services can tap into the experiences of all their customers to give everyone a sense of what’s going on. Nand Mulchandani, co-founder and CEO of SaaS startup ScaleXtreme calls it “the wisdom of the crowd.” Ideally, for example, his server-management service should be able to realize I’m running my servers at far too low a utilization rate compared with other customers, and it should be able tell me the ideal size instance to run and the most-effective software stack for fixing my problem.

As of Thursday, ScaleXtreme does just that. The company announced its first suite of analytics functions for configuration management, patch management, capacity and utilization, and monitoring. The analytics functions are fairly elementary off the bat, Mulchandani said — what side effects has the latest Windows (s msft) patch created, what are the best software stacks, do customers running Amazon Spot Instances save more than those using Reserved Instances, etc. — but they’ll get more advanced and more predictive in time.

That’s when things get really interesting and become the Holy Grail of crowdsourced analytics. Mulchandani thinks it will be about a year before ScaleXtreme is predicting problems, but it’s coming. “If machine A fails … the first question after you cleaned up the mess is ‘Wow, what other system of mine are going to fail?’,” he explained. If it puts the right models in place to identify failure patterns across its customer base, ScaleXtreme should be able to give users that level of foresight.

Among its peers, however, ScaleXtreme isn’t necessarily unique in its thinking. Cloudability, which monitors users’ spending across their various cloud services, has told me it’s also working on incorporating big data techniques into its service. The idea there is that the company could tell users how their cloud spending and choice of services compare with those of other similarly situated users. Another startup, Newvem, uses crowdsourced insights to recommend better choice for Amazon Web Services customers.

Nodeable (see disclosure), which launched last year as a Twitter-like app for letting administrators keep track of their systems, is undergoing a significant shift into a real-time analytics engine. It’s relying on Hadoop and Storm, a real-time processing front end, to analyze and alert customers to system events as they’re happening, perhaps even before.

There’s a saying in analytics that more data trumps better algorithms, and this definitely seems to be one of those cases. A single organization can only collect and analyze its own data, but that’s just a drop in the bucket of all the data generated by everyone trying to do the same thing. Especially with a new model such as cloud computing, there could be a real value in learning from what your peers are doing.

Disclosure: Nodeable is backed by True Ventures, a venture capital firm that is an investor in the parent company of this blog, Giga Omni Media. Om Malik, the founder of Giga Omni Media, is also a venture partner at True.

Image courtesy of Shutterstock user HuHu.