Red Sky Ops ML Engine Tunes Kubernetes HPA

Carbon Relay has announced some key enhancements of its flagship artificial intelligence for IT operations (AIOps) platform. The latest version of Red Sky Ops (v1.7) includes new dynamic resource tracking and pre-baked queries designed to allow users to understand resource utilization and optimize the Kubernetes Horizontal Pod Autoscaler (HPA).

The HPA automatically increases or decreases the number of Pods in a replication controller, deployment, replica set, or stateful set in response to the workload's CPU or memory consumption. It's implemented as a Kubernetes API resource and a controller. Red Sky Ops studies, replicates, and stress-tests Kubernetes apps, and then proactively deploys optimal configurations.

The platform's machine learning (ML) engine can now tune the Kubernetes HPA, ensuring that the optimized configurations are identified and implemented to handle anticipated and real-time traffic spikes without overprovisioning, the company says. Pod size and target utilization are constantly tested and optimized which ultimately results in better application performance and lower costs.

"Until now, managing and maintaining consistent and high application performance and reliability in Kubernetes environments has proved to be complicated," said Matt Provo, co-founder and CEO of Carbon Relay, in a statement, "but preparing for application scale introduces an entirely new level of complexity that can't be addressed by manual tuning. By applying the same machine learning principles that we use to identify optimal application configurations to the HPA, we can deliver a much more effective performance testing experience that ultimately leads to scalable and stable applications."

Boston-based Carbon Relay has carved out a "unique middle ground between machine and human intelligence, where they leverage the strengths of both for maximum effectiveness," the company says. Its mission is to build AI-powered software products "designed to help people, not replace them."

The latest release also comes with new support for string valued parameters. The new capabilities "broaden the set of things you can optimize for, provides more detail on failed configurations, and gives you more flexibility in how you view and filter experiment results," said Steph Rifai, Carbon Relay product manager, in a blog post.

Industry analyst firm Gartner has been credited with coining the term "AIOps" to define an emerging set of tools and platforms designed to augment IT functions, such as event correlation and analysis, anomaly detection, root cause analysis, and natural language processing. The firm estimates the size of the AIOps platform market at between $300 million and $500 million per year. Gartner analysts believe that, by 2023, 40% of DevOps teams will augment application and infrastructure monitoring tools with AIOps) platform capabilities (Market Guide for AIOps Platforms).   

About the Author

John K. Waters is the editor in chief of a number of sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS.  He can be reached at [email protected].