Latest

More Posts

Autonomous Optimization Across Your Entire Kubernetes Stack
By Marie JaksmanNov 10, 20265 min read

Autonomous Optimization Across Your Entire Kubernetes Stack

Autonomous Optimization of entire Kubernetes stack

Why Karpenter Will Not Consolidate Your Nodes
By Tania DuggalNov 5, 202613 min read

Why Karpenter Will Not Consolidate Your Nodes

When Karpenter leaves an underused node running, the events name the cause. Walk through four common blockers and the smallest diff that fixes each one.

How Do You Identify What Makes Your Service Expensive to Start?
By Tania DuggalOct 26, 202615 min read

How Do You Identify What Makes Your Service Expensive to Start?

Learn how to measure time to Ready and CPU-seconds until Ready, capture startup profiles with JFR and async-profiler, and map five common startup cost categories to sizing decisions.

Java Garbage Collection in Kubernetes: Choosing and Sizing the Right Collector
By Tania DuggalOct 21, 202617 min read

Java Garbage Collection in Kubernetes: Choosing and Sizing the Right Collector

The JVM picks a garbage collector from your pod's CPU and memory limits, and a small change can switch it silently. Learn how G1, ZGC, Shenandoah, Parallel, and Serial compare, and which one to run at each pod size.

Cost Problem with Overinflated Requests
By Tania DuggalOct 21, 202617 min read

Cost Problem with Overinflated Requests

JVM startup peaks push teams to set high CPU requests that last the pod's whole life. Learn how to measure the request gap and fix it safely with probes, boosts, and in-place resize.

Init Spike Handling: Right-Size JVM Pods for Boot and Run
By Ira ChernousOct 21, 20266 min read

Init Spike Handling: Right-Size JVM Pods for Boot and Run

Init Spike Handling gives JVM pods startup resources at boot, then resizes them in place for steady state. Cut boot-time waste with no restarts.

How to Fix Kubernetes Resource Policies That Optimize for the Wrong Phase
By Josh PalmerOct 19, 202619 min read

How to Fix Kubernetes Resource Policies That Optimize for the Wrong Phase

Most Kubernetes resource quota and LimitRange policies were written for steady state, so they throttle startup. Or they were written for startup, so they burn quota on idle reservations. Here's how to tell which one you have and how to fix it.

Kubecost Helm Chart: 2.x vs 3.x, Installation, and 6 Best Practices
By Tania DuggalOct 16, 202611 min read

Kubecost Helm Chart: 2.x vs 3.x, Installation, and 6 Best Practices

How to install Kubecost with Helm: the cost-analyzer and kubecost charts, 2.x vs 3.x, a step-by-step install, Helm vs manifests, and six best practices.

Cluster Autoscaler Helm Chart: How to Install and Configure It
By Tania DuggalOct 16, 202614 min read

Cluster Autoscaler Helm Chart: How to Install and Configure It

How to install and configure the Cluster Autoscaler Helm chart: how it decides to add and remove nodes, which chart to use, key values, tuning, and troubleshooting.

Kubernetes Deployment Strategy: Types and How to Choose
By Tania DuggalOct 16, 202614 min read

Kubernetes Deployment Strategy: Types and How to Choose

How Kubernetes deployment strategies work: the two built-in types, patterns that need a mesh or controller, their capacity costs, and how to match one to your workload.

Karpenter AWS: The Complete Guide to EKS Node Autoscaling
By Tania DuggalOct 15, 202615 min read

Karpenter AWS: The Complete Guide to EKS Node Autoscaling

A complete guide to Karpenter on AWS: how it works with EKS, how it compares to Cluster Autoscaler, a step-by-step tutorial, and best practices.