If you’re using Kubernetes’ Horizontal Pod Autoscaler (HPA) today, you know its shortcomings. Kubernetes can autoscale your applications based on metric thresholds such as CPU and memory utilization using the Kubernetes HPA resource. The goal of HPA is critical in ensuring that your application can handle the current demand to meet your SLOs and optimize the amount of resources that your application uses. In practice, most people find that the default HPA falls short of their needs.

We are building out Predictive Scaling functionality in our cloud-based platform that leverages Kubernetes HPA. We use the HPA API and add rich telemetry data and ML to model your applications’ behavior and give insight into when applications should be scaled out or in.

Read this white paper from our product engineering team: Overcome the limitations of Kubernetes’ Autoscaler (HPA) to achieve reliable, dynamic autoscaling

The white paper covers:

  • How HPA works
  • Biggest obstacles HPA users face today
  • How custom metrics adapters are used try to overcome limitations
  • How HPA can dynamically autoscale reliably and without expensive overprovisioning.

I invite you to partner with us today to see how our platform works in your environment. Our Early Access program provides full access and we’ll be adding new functionality for Continuous Rollouts and multi-cluster management over the next months.

Better yet, let’s meet and we’ll share what we are working on with you – and you can tell us what you want to achieve.