liqo
liqo copied to clipboard
kubectl run -it does not work with offloaded pods
What happened:
Trying to run a pod in a remote cluster through kubectl run -it ...
did not result in the shell correctly attached, and instead led to a pod in CrashLoopBackoff status.
What you expected to happen:
The shell should be attached correctly to the remote pod.
How to reproduce it (as minimally and precisely as possible):
- liqoctl offload namespace default --pod-offloading-strategy=Remote
- kubectl run ubuntu --image=ubuntu -n default -it --rm -- /bin/sh
- Observe that the shell is not attached correctly
Anything else we need to know?:
The kubelet emitted the following log line:
I1207 18:53:11.207826 1 klogv2.go:71] 404 request not found [uri=/attach/default/ubuntu/ubuntu?input=1&output=1&tty=1 vars=map[]]
Environment:
- Liqo version: 0.3.2-beta1
- Kubernetes version (use
kubectl version
): 1.21.0 - Cloud provider or hardware configuration: Kubeadm (on premise)
- Network plugin and version: Calico
- Install tools: liqoctl + helm
- Others: None