kubernetes: Error message: "failed to get azure cloud in GetVolumeLimits" in bare-metal setup after upgrade to 1.12.1
Is this a BUG REPORT or FEATURE REQUEST?:
Uncomment only one, leave it on its own line:
/kind bug
/kind feature
What happened: After upgrade of Kubernetes 1.11.3 to 1.12.1 via kubeadm in bare-metal setup I see these errors in kubelet logs:
Oct 8 12:27:50 node2 kubelet: E1008 12:27:50.766520 21094 azure_dd.go:147] failed to get azure cloud in GetVolumeLimits, plugin.host: node2.mydomain Oct 8 12:28:00 node2 kubelet: E1008 12:28:00.790762 21094 azure_dd.go:147] failed to get azure cloud in GetVolumeLimits, plugin.host: node2.mydomain etc…
What you expected to happen: Since this is a bare-metal setup I expect no messages concerning Azure cloud
How to reproduce it (as minimally and precisely as possible): Do an upgrade from 1.11.3 to 1.12.1
Anything else we need to know?: kubelet-1.12.1-2.x86_64.rpm was downloaded from https://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64
Environment:
- Kubernetes version (use
kubectl version
): Client Version: version.Info{Major:“1”, Minor:“12”, GitVersion:“v1.12.1”, GitCommit:“4ed3216f3ec431b140b1d899130a69fc671678f4”, GitTreeState:“clean”, BuildDate:“2018-10-05T16:46:06Z”, GoVersion:“go1.10.4”, Compiler:“gc”, Platform:“linux/amd64”} Server Version: version.Info{Major:“1”, Minor:“12”, GitVersion:“v1.12.1”, GitCommit:“4ed3216f3ec431b140b1d899130a69fc671678f4”, GitTreeState:“clean”, BuildDate:“2018-10-05T16:36:14Z”, GoVersion:“go1.10.4”, Compiler:“gc”, Platform:“linux/amd64”} - Cloud provider or hardware configuration: bare-metal
- OS (e.g. from /etc/os-release): Oracle Linux Server 7.4
- Kernel (e.g.
uname -a
): 4.9.62 - Install tools:
- Others:
About this issue
- Original URL
- State: closed
- Created 6 years ago
- Reactions: 14
- Comments: 42 (13 by maintainers)
I understand that the feature Node-specific Volume Limits has now reached beta in 1.12, but for those of us running Kubernetes on-premises, this log error is pretty useless. In fact, I would even want to disable this plugin alltogether, since it will never be used on-prem. Is there a way to stop the plugin from loading, or in the very least make sure the log does not appear? Right now, it is polluting the system log file quite a lot.
I second that. That was the key message of my bug report. That is we must not even instantiate cloud-specific plugins if no cloud provider is configured.
Same problems +1
disable
AttachVolumeLimit
in controller plane should work: “–feature-gates=AttachVolumeLimit=false”@D-PENG I have cherry picked the fix to v1.12
Update tho 1.12.2 resolve for me
Em qui, 1 de nov de 2018 9:22 PM, Abhilash Pallerlamudi < notifications@github.com escreveu:
I haven’t updated to 1.12.2 yet (still on 1.12.1) but just to confirm, disabling that feature works.
For those that are also running on CentOS and/or RHEL, just to the following:
Edit file
/etc/sysconfig/kubelet
on your master server (control plane), and add the following:KUBELET_EXTRA_ARGS=--feature-gates=AttachVolumeLimit=false
If you already have extra arguments in there, just add this one at the end of the line. Then restart kubelet on your master server:
sudo systemctl restart kubelet
It’s already fixed in v1.12.2 and v1.13.0. @gnufied do you know how to disable
AttachVolumeLimit
in kubelet in v1.12? Would add--feature-gates=AttachVolumeLimit=false
in kubelet command arguments work?