Open regaltopax opened 1 year ago
/triage accepted /assign @yangjunmyfm192085
Hi, @regaltopax, Please check whether the kubernetes cluster is configured with LimitRange?
Hi @yangjunmyfm192085 . I checked limits on the cluster, there is no limitrange applied on it
kubectl get limits -A No resources found
explicitly specifying resource limits along with the resource requests does make it work on AKS 1.25
Same thing happened to me in AKS. Turns out metrics server was already installed and I think the new deployment was somehow clashing with the existent one.
Hi,
I am running into this issue when running an AKS cluster with 1.24 and 1.25 versions.
Command used kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/download/metrics-server-helm-chart-3.8.3/components.yaml
This error is consistent no matter the version of metrics server used on the cluster. AKS 1.23 cluster doesnt have the issue. I also do not see this issue on EKS 1.24 either
Error The Deployment "metrics-server" is invalid:
I have 2 nodepools with the same vm size Standard_D8_v3, which has 8vcpu's and 32GiB Memory
kubectl get nodes output NAME CPU(cores) CPU% MEMORY(bytes) MEMORY% aks-pool1-xxx-vmss000000 110m 1% 4032Mi 14% aks-default-xxx-vmss000000 175m 2% 3599Mi 12%
AKS 1.24
My current workaround was to download the chart, remove the resource section, and run which works, but its not ideal since it does not have the necessary resource requirement
Happy to provide further information /kind support