Virtual Instruments wants AI to manage your data center

by Max Smolaks 4 September 2019

American software developer Virtual Instruments has launched a major update to VirtualWisdom, its software suite for AI-assisted data center management.

The company is a trailblazer in the emerging AIOps field, believing that server farms are best managed in partnership with algorithms that can sift though reams of logs and metrics to find ways to reduce IT infrastructure costs or identify issues that could cause downtime down the line.

“AIOps is an umbrella term and it can mean everything from doing pattern recognition in terms of IT operations, to applying various algorithms to the management of IT. Our approach is somewhat unique in that there’s application of purposeful algorithmic intelligence depending on the circumstances,” John Gentry, CTO of Virtual Instruments, told AI Business.

“At the core, the platform has a machine learning engine, and everything is based on understanding what is the normal behavior in the environment and being able to recognize anomalies.

“And then the application of AI in that context relates to very specific use cases: whether that’s looking at causality between trends that are happening in conjunction and understanding the relationships involved, or whether that’s running simulations to balance workloads and look for optimal placement of workloads in the environment.

“It really depends on what we’re solving for, based on our experience, with a combination of statistical analysis, heuristics, machine learning – all brought to bear against the problem.”

The company started out as a provider of infrastructure monitoring, but embraced machine learning technologies in 2014, and eventually built its own WisdomAI engine.

Among other things, the VirtualWisdom platform can forecast future consumption of resources like memory, compute and storage; it can help IT teams understand where specifically their applications reside and optimize application performance by moving workloads when appropriate. All of this functionality works across both on-premise infrastructure and public cloud.

Version 6.2 of VirtualWisdom adds brand new features like the Workload RightSizer, which can suggest the exact configuration of virtual machines based on the applications they are running, and even create a script that can be fed into change management software like ServiceNow to tweak configurations on-the-fly. Eventually, Virtual Instruments hopes to fully automate this process.

Workload RightSizer in action

Another new addition is the Drift Analyzer, which can warn when changes in application behavior are anomalous and deliver real-time alerts to restore application performance.

There are updates to predictive capacity management and proactive dashboards that support annotation, linking, and smart naming; and finally, WorkloadWisdom gets full integration with Dell EMC’s Isilon, the venerable family of enterprise storage appliances that has been a data center mainstay for almost 15 years.