👀 Views: 1329 💬 Answers: 1 📅 Created: 2025-06-14

kubernetes dns service deployment yaml

I'm trying to debug Quick question that's been bugging me - I'm working with an scenario where the DNS resolution for my Kubernetes service stops working after I update the deployment. Specifically, I have a service that routes traffic to a set of pods, and when I trigger a rolling update for the deployment, the service seems to temporarily lose its ability to resolve the pod IP addresses. This results in a brief outage for users trying to access the service. Here’s a simplified version of my service and deployment configuration: ```yaml apiVersion: apps/v1 kind: Deployment metadata: name: my-app spec: replicas: 3 selector: matchLabels: app: my-app template: metadata: labels: app: my-app spec: containers: - name: my-container image: my-image:v1.0 ports: - containerPort: 80 --- apiVersion: v1 kind: Service metadata: name: my-service spec: selector: app: my-app ports: - protocol: TCP port: 80 targetPort: 80 type: ClusterIP ``` When I run `kubectl rollout restart deployment my-app`, the deployment updates the pods as expected, but I notice in my logs that for a few seconds, DNS queries to `my-service` either time out or return `NXDOMAIN`. I’ve tried increasing the `dnsPolicy` to `ClusterFirstWithHostNet`, but it didn’t resolve the scenario. I also confirmed that the pods are in a Running state and the service endpoints look correct using `kubectl get endpoints my-service`. I also checked the CoreDNS logs and found entries like `plugin/rewrite: rewrite of my-service failed` which seems to indicate issues with the DNS rewrites during the update. Is there a best practice to handle this situation, or perhaps a configuration I’m missing that could help maintain DNS resolution during deployment updates? The Kubernetes version I'm using is v1.30. Any insights would be greatly appreciated. The stack includes Yaml and several other technologies. I'm coming from a different tech stack and learning Yaml.