kubectl get nodes
Should provide an output like:
NAME STATUS ROLES AGE VERSION
ip-172-31-94-139.eu-central-1.compute.internal Ready control-plane,master 2d10h v1.21.3
ip-172-31-94-250.eu-central-1.compute.internal Ready <none> 2d10h v1.21.3
ip-172-31-95-121.eu-central-1.compute.internal Ready <none> 2d10h v1.21.3
ip-172-31-95-122.eu-central-1.compute.internal Ready control-plane,master 2d10h v1.21.3
ip-172-31-96-137.eu-central-1.compute.internal Ready control-plane,master 2d10h v1.21.3
ip-172-31-96-184.eu-central-1.compute.internal Ready <none> 2d10h v1.21.3
kubectl get pod -A
If there are any pods in Pending state - check if the reason is insufficient resources and scale the MachineDeployment resources accordingly.
You can scale MachineDeployment resource(s) with following command: kubectl scale machinedeployment -n kube-system <md-name> --replicas=2.
If there are some failing pods, investigate the specific logs.
kubectl get kustomizations.kustomize.toolkit.fluxcd.io -A
Should provide an output like:
NAMESPACE NAME READY STATUS AGE
flux-system flux-system True Applied revision: main/1d6cdc2b442c11c30ee005430b25b27f668ed5f9 2d9h
kubermatic sops-kustomization True Applied revision: main/1d6cdc2b442c11c30ee005430b25b27f668ed5f9 2d9h
kubectl get helmreleases.helm.toolkit.fluxcd.io -A
Should provide an output like:
NAMESPACE NAME READY STATUS AGE
iap iap True Release reconciliation succeeded 2d9h
kube-system s3-exporter True Release reconciliation succeeded 2d9h
minio minio True Release reconciliation succeeded 41h
monitoring alertmanager True Release reconciliation succeeded 2d9h
monitoring blackbox-exporter True Release reconciliation succeeded 2d9h
monitoring grafana True Release reconciliation succeeded 2d9h
monitoring karma True Release reconciliation succeeded 2d9h
monitoring kube-state-metrics True Release reconciliation succeeded 2d9h
monitoring node-exporter True Release reconciliation succeeded 2d9h
monitoring prometheus True Release reconciliation succeeded