Kubeadm: kubelet.service: Service has more than one ExecStart= setting

[joshuacox]

What keywords did you search in kubeadm issues before filing this one?

ExecStart

Is this a BUG REPORT or FEATURE REQUEST?

/kind bug

Versions

kubeadm version (use kubeadm version):
v1.12.2

Environment:

• Kubernetes version (use kubectl version):

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"}

• Cloud provider or hardware configuration:
  bare metal hosts running KVM
• OS (e.g. from /etc/os-release):
  ubuntu on KVM
• Kernel (e.g. uname -a):

uname -a
Linux traqqymaster1 4.4.0-131-generic #157-Ubuntu SMP Thu Jul 12 15:51:36 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux

• Others:
  all scripted with kubash

What happened?

systemctl status kubelet
● kubelet.service - kubelet: The Kubernetes Node Agent
   Loaded: error (Reason: Invalid argument)
  Drop-In: /etc/systemd/system/kubelet.service.d
           └─10-kubeadm.conf, 20-etcd-service-manager.conf
   Active: inactive (dead) (Result: exit-code) since Wed 2018-10-31 20:36:15 UTC; 17min ago
     Docs: https://kubernetes.io/docs/home/
 Main PID: 2538 (code=exited, status=255)

Oct 31 20:36:05 traqqymaster1 systemd[1]: kubelet.service: Main process exited, code=exited, status=255/n/a
Oct 31 20:36:05 traqqymaster1 systemd[1]: kubelet.service: Unit entered failed state.
Oct 31 20:36:05 traqqymaster1 systemd[1]: kubelet.service: Failed with result 'exit-code'.
Oct 31 20:36:15 traqqymaster1 systemd[1]: Stopped kubelet: The Kubernetes Node Agent.
Oct 31 20:36:24 traqqymaster1 systemd[1]: kubelet.service: Service has more than one ExecStart= setting, which is only allowed for Type=oneshot services. Refusing.
Oct 31 20:37:33 traqqymaster1 systemd[1]: kubelet.service: Service has more than one ExecStart= setting, which is only allowed for Type=oneshot services. Refusing.

What you expected to happen?

kubelet to be running

How to reproduce it (as minimally and precisely as possible)?

kubash -n testcluster auto

Anything else we need to know?

I am downloading this file in my script:

https://raw.githubusercontent.com/kubernetes/kubernetes/v1.12.2/build/debs/10-kubeadm.conf

Which has two ExecStart lines in it.

[yagonobre]

Please create a issue in https://github.com/kubash/kubash

[joshuacox]

@yagonobre I am the author of kubash, and the issue is with this file:

https://raw.githubusercontent.com/kubernetes/kubernetes/v1.12.2/build/debs/10-kubeadm.conf

not a part of kubash

[yagonobre]

@joshuacox you opened a pr in your own repository.
Also you have any ExecStart on 20-etcd-service-manager.conf? I guess that a empty ExecStart is not a problem.

[neolit123]

an empty ExecStart= clears settings of a previous ExecStart.

[neolit123]

your 20-etcd-service-manager.conf should look like this:
https://kubernetes.io/docs/setup/independent/setup-ha-etcd-with-kubeadm/#setting-up-the-cluster

20- implies higher precedence and it also contains ExecStart=

[joshuacox]

@neolit123 @yagonobre indeed the initial ExecStart is a clear, and the 20-etcd-service-manager.conf has the same thing in it:

https://kubernetes.io/docs/setup/independent/setup-ha-etcd-with-kubeadm/

and my old template file did not have the empty ExecStart line which was conflicting!