CoreDNS error in AKS - can't reach applicaiton

Question

We have a Kubernetes cluster that suddenly stopped working. The Resource Health tab says the information below. I have rebooted coreDNS but it didn't resolve the issue. The application is completely unreachable at this time. Anything to try?

Degraded (Customer Initiated) : CoreDNS became unavailable (Customer Initiated)

---

At Monday, August 21, 2023 at 3:43:33 AM MDT, the Azure monitoring system received the following information regarding your Azure Kubernetes Service (AKS):

We've temporarily lost connection to Core DNS. If the AKS cluster is running as expected, the status of the resource will change to Available after a few minutes.

Recommended Steps

• Check back here for status updates. AKS is trying to recover this issue
• For coreDNS Troubleshooting please check the CoreDNS Troubleshooting documentation.
• Problems with CoreDNS may be a symptom of wider issues with you cluster, visit Diagnose and solve problems to further diagnose and troubleshoot the issue.
• If this error persists and you are having issues interacting with your cluster contact support

Answer 1

We resolved this but were forced to upgrade our cluster to get it resolved. I'm not entirely sure what happened to the DNS pods. They were healthy. We then had to rebuild our ingress to match the new spec.

Answer 2

Hello @joe white If you see the Error message below on the portal.

• We've temporarily lost connection to Core DNS. If the AKS cluster is running as expected, the status of the resource will change to Available after a few minutes.

• Then, you do not need to do anything as the Coredns will resolve itself, just give it something for it to change it status, whether it is in a CrashLoopBackOff.

Since it is not what you manage.