Couldn't get resource list for metrics.k8s.io; connect: no route to host #3288
Comments
|
Initially we used network:
plugin: calicoNow we've switched to default network:
plugin: flannel
options: {}and redeployed this testing cluster: "network": {
"plugin": "flannel",
"options": {
"flannel_backend_port": "8472",
"flannel_backend_type": "vxlan",
"flannel_backend_vni": "1"
},Somebody recommended adding more resources to - apiGroups:
- ""
resources:
- pods
- nodes
- nodes/stats # added
- namespaces # added
- configmaps # added
verbs:
- get
- list
- watchbut I'm afraid that didn't help. |
|
Here are some other steps I've made without much success.
|
|
It was working with |
|
Finally I has started working bot with Calico and Flannel, surely not at the same time. 😄 |
|
@manuelbuil, interestingly today we hit this issue again with However upon investigation we didn't find any interfaces with cali*. Is there anything we should look closely? |
|
This looks like a problem with Calico deployment. NetworkPluginNotReady message:docker: network plugin is not ready: cni config uninitializedOne thing concerns me. AFAIK file |
|
Let me provide another piece of details related to issues with Calico
|
|
Could you get the logs from the image that is crashing please? I wonder if there is something like apparmor or selinux making it impossible for Calico to write binaries in |
|
@manuelbuil, this is CentOS 7.9 and I'm positive it has SELinux disabled. |
|
This repository uses an automated workflow to automatically label issues which have not had any activity (commit/comment/label) for 60 days. This helps us manage the community issues better. If the issue is still relevant, please add a comment to the issue so the workflow can remove the label and we know it is still valid. If it is no longer relevant (or possibly fixed in the latest release), the workflow will automatically close the issue in 14 days. Thank you for your contributions. |
|
I still facing this issue, do we have a fix? |
|
This repository uses an automated workflow to automatically label issues which have not had any activity (commit/comment/label) for 60 days. This helps us manage the community issues better. If the issue is still relevant, please add a comment to the issue so the workflow can remove the label and we know it is still valid. If it is no longer relevant (or possibly fixed in the latest release), the workflow will automatically close the issue in 14 days. Thank you for your contributions. |
RKE version:
Kubernetes version:
As reported by
kubectl get nodesDocker version: (
docker version,docker infopreferred)As reported by
kubectl get nodesOperating system and kernel:
uname: 3.10.0-1160.el7.x86_64
Created anew with
rke upon four VMs (QEMU/KVM) CentOS 7.9.The testing environment is comprised of three nodes with
controlplane,etcdand oneworker.Type/provider of hosts: (VirtualBox/Bare-metal/AWS/GCE/DO)
QEMU/KVM (kernel 5.10.152)
Steps to Reproduce:
Frankly hard to report any specific as the setup was very straightforward
rke configOfficial instructions was followed: https://rke.docs.rancher.com/installation. Certificates were not customized, used default options.
Most of answers were default. Only calico was chosen instead of default flannel.
rke upAll images were pulled from Docker Hub. These nodes had been fresh until I started running
rke upon them.Results:
Initially we got embarrassed by behaviour of
kubectltool which caused the same four error lines complaining about metrics-serverWe started looking into the issue and our investigation led us to several other problems.
metrics-server-xxxxxxxxxx-xxxxxis in status CrashLoopBackOffIts logs show these lines below.
k logs metrics-server-xxxxxxxxxx-xxxxxWhile
calico-nodeitself looks healthy.calico-kube-controllersare full of these messages.Interestingly, despite these issues, this cluster is still capable of running workload.
I'm looking for a solution. Meanwhile I've found something resembling our issues:
The text was updated successfully, but these errors were encountered: