elotl / kip

Virtual-kubelet provider running pods in cloud instances
Apache License 2.0
223 stars 14 forks source link

issues deploying pods to kip provider in minikube #248

Open GuopingJia opened 2 years ago

GuopingJia commented 2 years ago

I've managed to set up the kip provider to my minikube cluster by following up your scripts in deploy/manifests/kip/overlays/minikube/ (with provided aws credentials in provider.yaml):

$ k get nodes NAME STATUS ROLES AGE VERSION kip-provider-0 Ready agent 19h minikube Ready control-plane,master 175d v1.20.2

Then, I've started deploying a testing nginx pod to the kip provider to its 'default' (by adding the expected nodeSelector/Toleration in the deployment manifest file). I can see the aws ec2 instance being created in aws. And the instance has been finally showing in running state. However, from my cluster, the pod keeps showing in Pending state.

$ k get pods,deploy NAME READY STATUS RESTARTS AGE pod/nginx-kip-65b77ccc57-q2wbk 0/1 Pending 0 94m

NAME READY UP-TO-DATE AVAILABLE AGE deployment.apps/nginx-kip 0/1 1 0 94m

Below error is showing up when try to get the pod logs:

Error from server (InternalError): Internal error occurred: error getting container logs?): finding logs for "default_nginx-kip-65b77ccc57-q2wbk"/"nginx": Error getting older log from registry default_nginx-kip-65b77ccc57-q2wbk: Could not get log default_nginx-kip-65b77ccc57-q2wbk/nginx from storage: Key not found in store