Kubespray: containerd not using systemd as cgroup driver

[matjazp]

Environment:

• baremetal

• OS:
  Ubuntu 18.04.3 LTS

• Version of Ansible (ansible --version):
  ansible 2.7.12

Kubespray version (commit) (git rev-parse --short HEAD):
HEAD

Network plugin used:
calico

I've been trying to use containerd with systemd as a cgroup driver, as I believe that on Ubuntu and all other systemd systems it should be used as a default (see #5134 when using Docker). When using containerd and forcing usage of systemd as a cgroup driver with:

containerd_use_systemd_cgroup: true

I found that containerd still uses the default, which is not systemd:

$ sudo crictl info | grep system
   “systemdCgroup”: false,

Problem is we're not setting that in /etc/containerd/config.tomlas we should according to https://kubernetes.io/docs/setup/production-environment/container-runtimes/#systemd.

[rjbez17]

I suspect you'll see this in your logs:

error="invalid plugin config: `systemd_cgroup` only works for runtime io.containerd.runtime.v1.linux

This may or may not work:

[plugins]
  [plugins.cri]
    [plugins.cri.containerd.runtimes]
      [plugins.cri.containerd.runtimes.runc]
         runtime_type = "io.containerd.runc.v1"
         [plugins.cri.containerd.runtimes.runc.options]
           SystemdCgroup = true

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale