CORE DNS PODS IN KUBE-SYSTEM NOT WORKING
I have a fargate class in EKS but my Core DNS pods are in a pending state. How do I go about the issue?
- Le plus récent
- Le plus de votes
- La plupart des commentaires
Hello Joash,
By default, CoreDNS deployment manifest will have an annotation eks.amazonaws.com/compute-type : ec2 which will make sure that coredns pods get scheduled only in nodes with EC2 type. As you are trying to run the coredns pods in Fargate nodes, you will have to remove this annotation.
To do this manually, use the following kubectl command line interface (CLI) commands, in sequential order, after you create the Amazon EKS cluster:
kubectl patch deployment coredns -n kube-system --type=json -p='[{"op": "remove", "path": "/spec/template/metadata/annotations", "value": "eks.amazonaws.com/compute-type"}]'
kubectl rollout restart -n kube-system deployment coredns
Also, please make sure that your cluster has a fargate profile available for the kube-system namespace.
If you are still having issues after checking the above suggestions, please share the coredns pod describe output.
Hope this is helpful!
If any pods are stuck in the Pending state, confirm that your node instance type is listed in limits.go
Contenus pertinents
- demandé il y a un an
- demandé il y a un an
- Pourquoi le plug-in CNI de mon VPC ne parvient-il pas à atteindre le serveur d'API dans Amazon EKS ?
AWS OFFICIELA mis à jour il y a 2 ans - AWS OFFICIELA mis à jour il y a 7 mois
- AWS OFFICIELA mis à jour il y a un an
- AWS OFFICIELA mis à jour il y a 4 mois
It was helpful. Thank you Do I need different fargate profiles for my frontend and backend? How can I deploy my backend and frontend in EKS?
Hey Joash, you can create one fargate profile with a specific namespace and run both your frontend and backend workloads in that namespace to run them on EKS fargate.
If we configure cordons this way, then how can we spin up both EC2 and Fargate load together ?