Merge pull request #291727 from papadeltasierra/pauldsmith/537853847

Troubleshooting issues with DNS in Nexus Network Fabric

Author: v-ccolin

articles/operator-nexus/TOC.yml

@@ -330,6 +330,8 @@
           href: troubleshoot-memory-limits.md
         - name: Troubleshoot LACP Bonding
           href: troubleshoot-lacp-bonding.md
+        - name: Troubleshoot DNS Issues
+          href: troubleshoot-dns-issues.md
         - name: Troubleshoot NAKS Cluster Node Packet Loss
           href: troubleshoot-packet-loss.md
         - name: Troubleshoot TWAMP (UDP) not working

articles/operator-nexus/troubleshoot-dns-issues.md

@@ -0,0 +1,56 @@
+---
+title: "Azure Operator Nexus: DNS Issues"
+description: Learn how to troubleshoot cluster DNS issues.
+author: papadeltasierra
+ms.author: pauldsmith
+ms.service: azure-operator-nexus
+ms.custom: azure-operator-nexus
+ms.topic: troubleshooting
+ms.date: 12/10/2024
+# ms.custom: template-include
+---
+
+# Troubleshoot Nexus DNS Issues
+
+NNF (Nexus Network Fabric) provides a bridge between Nexus resources hosted by a Kubernetes
+cluster running on Azure VMs (Virtual Machines) and Azure, accessing Azure resources via their
+domain names. However a DNS (Domain Name System) error in NNF can mean that Azure resources
+can't be contacted which impacts deployment or management of Nexus resources.
+
+The DNS proxy that causes this error is an [Envoy DNS Proxy](https://www.envoyproxy.io/docs/envoy/latest/)
+running via a Kubernetes deployment in either an infrastructure or tenant Kubernetes cluster.
+The precise location of the DNS proxy is determined when the customer
+deploys their NAKS (Nexus Azure Kubernetes Service) cluster or during some other
+deployment. 
+
+## Diagnosis
+
+* Deployment or management of remote Nexus resources fails with "DeploymentFailed."
+* Azure portal shows no errors being generated for the Azure resources that are unreachable; there are no errors because the failing operations aren't reaching the Azure resources at all.
+
+## Mitigation steps
+
+### Trigger a DNS cache refresh for the NNF Workload Proxy
+
+- Identify the Infrastructure or Tenant Kubernetes Cluster on which the DNS proxy is running from the initial configuration and deployment process
+- Log in to the Kubernetes cluster
+  - Using the Azure portal, find your cluster
+  - From the _Overview_ blade, click the _Connect_ command (between _Refresh_ and _Delete_)
+  - Follow the instructions from the resulting pop-up window that explain how to connect to the Kubernetes cluster
+- Identify the DNS proxy deployment using this command
+  ```bash
+  $ kubectl get deployments --all-namespaces=true | grep envoy
+  ```
+- Restart the deployment, which causes the DNS caching to be reset, using this command:
+  ```bash
+  kubectl rollout restart deployment <your-envoy-deployment-name> --namespace <namespace-where-envoy-pod-exists>
+  ```
+
+## Verification
+
+After the DNS cache is refreshed, create or manage operations are successful.
+
+## Related content
+
+- If you still have questions, contact [Azure support](https://portal.azure.com/?#blade/Microsoft_Azure_Support/HelpAndSupportBlade).
+- For more information about support plans, see [Azure support plans](https://azure.microsoft.com/support/plans/response/).