kubeasz版本3.1.0，执行 ezctl setup 集群 05，安装node 这一步时报错 'DOCKER_VER' is undifined

[whenandwhy]

What would you like to be added?

整个集群六台服务器，执行setup 05 master没问题worker node报错 （TASK [kube-node : 创建kubelet的配置文件] ** fatal: [XX.XX.XX.94]: FAILED! => {"changed": false, "msg": "AnsibleUndefinedVariable: {%- if CONTAINER_RUNTIME == 'containerd' -%} systemd {%- else -%} {%- if DOCKER_VER|float >= 20.10 -%} systemd {%- else -%} cgroupfs {%- endif -%} {%- endif -%}: 'DOCKER_VER' is undefined"} fatal: [XX.XX.XX.95]: FAILED! => {"changed": false, "msg": "AnsibleUndefinedVariable: {%- if CONTAINER_RUNTIME == 'containerd' -%} systemd {%- else -%} {%- if DOCKER_VER|float >= 20.10 -%} systemd {%- else -%} cgroupfs {%- endif -%} {%- endif -%}: 'DOCKER_VER' is undefined"} fatal: [XX.XX.XX.96]: FAILED! => {"changed": false, "msg": "AnsibleUndefinedVariable: {%- if CONTAINER_RUNTIME == 'containerd' -%} systemd {%- else -%} {%- if DOCKER_VER|float >= 20.10 -%} systemd {%- else -%} cgroupfs {%- endif -%} {%- endif -%}: 'DOCKER_VER' is undefined"}）

其中hosts文件配置如下

'etcd' cluster should have odd member(s) (1,3,5,...)

[etcd] XX.XX.XX.91 XX.XX.XX.92 XX.XX.XX.93

master node(s)

[kube_master] XX.XX.XX.91 XX.XX.XX.92 XX.XX.XX.93

work node(s)

[kube_node] XX.XX.XX.91 XX.XX.XX.92 XX.XX.XX.93 XX.XX.XX.94 XX.XX.XX.95 XX.XX.XX.96

Why is this needed?

求高手解答

[gjmzj]

之前老版本kubelet使用docker 运行时 cgroups 支持 systemd 不好，目前主流都使用containerd 运行时， 把 /root/repo/kubeasz/roles/kube-node/vars/main.yml 变量改成如下就好

CGROUP_DRIVER: "systemd"

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.

[whenandwhy]

之前老版本kubelet使用docker 运行时 cgroups 支持 systemd 不好，目前主流都使用containerd 运行时， 把 /root/repo/kubeasz/roles/kube-node/vars/main.yml 变量改成如下就好

CGROUP_DRIVER: "systemd"

已解决，感谢高手解答