问题
I setup a Kubernetes cluster with four EC2 instances using kubeadm
. The Kubernetes cluster works fine, but failed when I try to create a PersistentVolumeClaim
.
First I created a StorageClass
with following YAML which works fine.
kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
name: generic
annotations:
storageclass.kubernetes.io/is-default-class: "true"
provisioner: kubernetes.io/aws-ebs
parameters:
type: gp2
encrypted: "false"
Then I try to create a PersistentVolumeClaim
with the following YAML:
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: slow
labels:
app: nginx
spec:
storageClassName: generic
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 3Gi
Which got an error message of:
Failed to provision volume with StorageClass "generic": Failed to get AWS Cloud Provider. GetCloudProvider returned <nil> instead
By the way, it seems no PersistentVolume
was ever created.
kubectl get pv
No resources found.
Any ideas how to solve this?
Thanks.
回答1:
This is because you didn't use the option --cloud-provider=aws
while running kubeadm init
kubeadm --cloud-provider=aws init ...
Alternatively, you should put the cloud provider info ( cloudProvider ) into your config file and it will add the cloud provider flag to api and controller manager during initialization :
apiVersion: kubeadm.k8s.io/v1alpha1
kind: MasterConfiguration
etcd:
endpoints:
- "http://my-etcd1"
- "http://my-etcd2"
- "http://my-etcd3"
cloudProvider: aws
apiServerExtraArgs:
apiserver-count: 3
apiServerCertSANs:
- "my-master-hostname1"
- "my-master-hostname2"
- "my-master-hostname3"
- "10.10.0.50"
- "10.10.0.51"
- "10.10.0.52"
- "127.0.0.1"
token: "my-kube-token"
tokenTTL: "0"
More info : https://medium.com/jane-ai-engineering-blog/kubernetes-on-aws-6281e3a830fe
来源:https://stackoverflow.com/questions/57876648/kubernetes-failed-to-create-persistentvolumeclaim-on-aws-ebs