Retrying Bicep Resource Deployments with the @retryOn Decorator
Use Bicep’s @retryOn decorator to retry resource deployments for selected transient Azure errors, with a practical role assignment example and the limits to keep in mind.
#Troubleshooting 7 articles
Articles tagged with "Troubleshooting"
Testing AKS Node Auto Provisioning: When Our Managed Disk Policy Blocked Every New Node
While evaluating a move from the cluster autoscaler to Node Auto Provisioning (NAP) on AKS, no nodes were created. NAP quietly fell back from ephemeral OS disks to managed disks, and our Azure Policy denied them.
AKS Desktop: A Hands-On Tour of Microsoft's New Desktop App for Kubernetes
What AKS desktop actually does — Projects, guided deployments without YAML, a natural-language AI troubleshooting assistant, and a built-in plugin catalog — walked through as a demo you can reproduce on your own cluster.
Azure SQL Error 9001: The Log for Database Is Not Available
A real-world look at SQL error 9001 in Azure SQL diagnostic logs — what it means, why it happens during planned maintenance reconfigurations, and how to make your application resilient against it.
When Good Policies Break Monitoring: How an Azure Policy Silently Broke Our AKS Pipeline
An AKS incident from our deployment pipeline: Azure Monitor addons moved to a cluster extension-based backend — and a sensible Azure Policy that blocked extensions suddenly broke our AKS provisioning with no warning.
Why You Should Never Lock AKS-Managed Resources: A Volume Outage Story
A real incident from our AKS platform: resource locks on AKS-managed Azure disks silently broke the CSI driver during a node drain, leaving pods stuck without volumes. Here's what went wrong and what we should have done instead.
Rescuing an AKS Namespace Stuck in Terminating State
A real-world story: how a nightly cleanup pipeline got blocked by a namespace stuck in Terminating, and the kubectl proxy + finalizer trick that finally got it unstuck.