Managing applications running in Kubernetes can be far more complex and time-consuming than most DevOps, networking and IT professionals expect. The platform’s flexibility is both its strength and weakness. Kubernetes allows experts to tune it to support their organization’s needs, yet it can be so tough that many teams find it frustrating to the point of being unmanageable. Teams are often forced to opt to dramatically overprovision compute and storage resources to ensure application performance, running up unsustainable costs.
To address this complexity, Carbon Relay created Red Sky Ops, an AIOps platform for deploying, scaling and managing containerized applications in Kubernetes environments. It uses machine learning to automatically determine the optimal configuration for apps running in Kubernetes, eliminating the need for ineffective manual optimization.
Using ML-powered experimentation, Red Sky Ops explores the application parameter space, resulting in configurations that both deploy reliably and perform optimally—a nearly impossible task for even the most capable DevOps teams to undertake by hand. We also created Red Sky Ops to learn over time, allowing the platform to become even more efficient over time.
Now, Carbon Relay is collaborating with the IBM Cloud Kubernetes Service, a complete managed container service, to tackle the Kubernetes complexity challenge head-on. I’ve worked with IBM’s Chris Rosen, program director, offering management of the IBM Kubernetes Service, to describe in detail the work we’re doing together to help deliver on the vision of Kubernetes.
In Turning a Glimpse of Kubernetes’ Future into Reality, Chris and I describe the collaboration between IBM and Carbon Relay, and how we’re providing enterprises with new and effective ways to use Kubernetes to achieve their business goals—reliably, efficiently, and flexibly.