K3s quickstart instructions not working (cni plugin not initialized)

Open Source Calico Help Kubernetes
1

Hello Community

The instructions on Quickstart for Calico on K3s seem to be incorrect.
When I copy the command with

INSTALL_K3S_EXEC="--flannel-backend=none

The pods remain in Pending mode and in the syslog I see

Dec 14 08:21:01 bntuu k3s[6067]: E1214 08:21:01.454319    6067 pod_workers.go:965] "Error syncing pod, skipping" err="network is not ready: container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized" pod="kube-system/metrics-server-5c8978b444-6zxgl" podUID=69d5e72d-45fa-4b76-bdfd-742fd4b79b3a

When I remove the ‘–flannel-backend=none’ the status of the pods becomes ‘Running’
But the syslog is still flooded with messages about cni

Dec 14 08:21:17 bntuu k3s[6067]: E1214 08:21:17.452641    6067 pod_workers.go:965] "Error syncing pod, skipping" err="network is not ready: container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized" pod="kube-system/coredns-597584b69b-z7j5r" podUID=920118f1-6400-4701-8f5b-c3f56f3c8bc7
Dec 14 08:21:17 bntuu k3s[6067]: E1214 08:21:17.452703    6067 pod_workers.go:965] "Error syncing pod, skipping" err="network is not ready: container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized" pod="kube-system/metrics-server-5c8978b444-6zxgl" podUID=69d5e72d-45fa-4b76-bdfd-742fd4b79b3a

How can this be resolved?

Regards Hans

2

You need to set up the cluster with no CNI plugin (--flannel-backend=none) and then follow the rest of the doc to install Calico.

A cluster with no CNI plugin is expected to be NotReady, and pods will not be scheduled until you have installed a CNI. This is expected.

3

The --flannel-backend=none disables the default flannel CNI plugin, which is why your cluster prompts that error. It will be resolved if you follow the tutorial and install the CNI.

Here are some resources that explain this problem in detail:
Instruqt ← hands-on lab environment