Azure

Azure Chaos Studio is a managed chaos engineering service that allows you to introduce controlled faults into your Azure environment to validate resilience and robustness. In this post, you’ll learn how to use Chaos Studio to run experiments on Azure Kubernetes Service (AKS) clusters. Prerequisites An existing AKS cluster Azure CLI installed and logged in kubectl configured AKS cluster must be running in a supported region Chaos Studio enabled in your subscription Register Required Resource Providers az provider register --namespace Microsoft.Chaos az provider register --namespace Microsoft.ContainerService Enable Chaos Studio on Your AKS Cluster az chaos target create --resource-id $(az aks show -g <resource-group> -n <aks-name> --query id -o tsv) \ --target-type Microsoft-AzureKubernetesServiceCluster \ --location <region> You can also enable Chaos Studio from the Azure Portal under your AKS resource > Chaos Studio > Enable. ...

The Node Autoprovisioner is a powerful open-source component that enables Kubernetes to automatically provision (and deprovision) nodes based on pod scheduling needs. It’s similar in intent to the Cluster Autoscaler, but it allows for greater flexibility by dynamically creating new node pools with custom configurations — including taints, labels, and instance types. This guide will walk you through deploying the Node Autoprovisioner on Azure Kubernetes Service (AKS). Prerequisites An existing AKS cluster (version >= 1.26 recommended) Azure CLI installed and logged in Helm installed Cluster-admin access (kubectl configured) The AKS cluster must use VMSS (Virtual Machine Scale Sets) Enable Managed Identity on the cluster Enable AKS Preview Features (if needed) Node Autoprovisioner requires enabling the AKS Preview CLI extension: ...