By Zesty Team

Read More >

We’re excited to launch Pod Rightsizing, the newest addition to Zesty Kompass, our Kubernetes optimization platform. This solution adds Vertical Pod Autoscaler (VPA) support, working alongside our existing Horizontal Pod Autoscaler (HPA) capabilities to automatically scale container resources based on real-time and historical usage.

With Pod Rightsizing, engineering teams can reduce costs and improve performance without manual intervention. By combining HPA and VPA on a single platform, we’re allowing teams to apply multiple-dimensional automation tools under the same platform, so they can gain deeper control over resource usage, boost performance, and maximize cost reduction.

Vertical Scaling: Powerful but Painful

Manually managing vertical scaling in Kubernetes is time-consuming and error-prone. Engineers must constantly monitor resource usage and manually adjust CPU and memory requests for each container.

Without automation, it’s easy to get it wrong, leading to CPU throttling if requests are too low or out-of-memory (OOM) errors that crash applications. On the flip side, overprovisioning drives up cloud costs without delivering any real performance benefits.

Kubernetes does offer a native Vertical Pod Autoscaler, but it comes with some limitations. It relies on historical data and does not respond to demand fluctuations in real time. It does not monitor OOM and CPU throttling, which can impact performance. Adjusting it for different workloads also takes time and increases the risk of misconfigurations. Most crucially, native VPA doesn’t integrate well with HPA, which is used by over half of Kubernetes users.

Optimal Resource Allocation with Pod Rightsizing

Our new Pod Rightsizing solution automatically adjusts CPU and memory requests based on real-time and historical data, ensuring that each container receives exactly the resources it needs to perform optimally. Through this automation process, Kompass reduces waste from static resource requests and avoids conflicts between VPA and HPA. These capabilities are key for multi-container pods, including sidecars, adapter containers, or tightly coupled helper containers.

Insights and recommendations at the workload level

By optimizing resources at the container level, we allow better precision in resource allocation, which is vital for maintaining system stability. By continuously tuning CPU and memory to fit the needs of each container, Kompass Pod Rightsizing prevents incidents of CPU throttling and monitors out-of-memory (OOM) incidents, ensuring that applications remain responsive and stable, even under fluctuating workloads.

With more precise and automated resource allocation, teams can align resources closely with actual usage, instead of maintaining excess capacity just to be on the safe side, thus eliminating cloud waste.

These accurate recommendations and automation capabilities also reduce the operational overhead associated with manual monitoring, allowing teams to be more efficient.

CPU and RAM utilization

Integrating Pod Rightsizing into the Zesty Kompass Suite

Pod Rightsizing is joining a vast suite of optimization solutions already available on Zesty Kompass, complementing them to form a comprehensive platform for managing K8s infrastructure. It works seamlessly with Horizontal Pod Autoscaler (HPA) to provide both vertical and horizontal resource adjustments, ensuring precise allocation tailored to the real-time needs of each application.

Pod Rightsizing also forms a robust resource optimization strategy at every layer of your K8s infrastructure. Solutions such as Headroom Reduction and PVC Autoscaling ensure resources are optimally used without waste. Additionally, our Spot Automation solution safely maximizes the use of Spot instances, unlocking cost savings without the risk of downtime.

Together, these technologies create a unified, holistic approach to Kubernetes optimization. Zesty Kompass allows engineering teams to manage scaling, optimize costs, and tune performance, leading to more resilient and cost-effective Kubernetes environments.

Want to learn more about Pod Rightsizing and how it can benefit your Kubernetes setup? Book a demo and see it firsthand.