Container Runtimes

You need to install a container runtime into each node in the cluster so that Pods can run there. This page outlines what is involved and describes related tasks for setting up nodes.

Kubernetes 1.28 requires that you use a runtime that conforms with the Container Runtime Interface (CRI).

See CRI version support for more information.

This page provides an outline of how to use several common container runtimes with Kubernetes.

• containerd
• CRI-O
• Docker Engine
• Mirantis Container Runtime

Install and configure prerequisites

The following steps apply common settings for Kubernetes nodes on Linux.

You can skip a particular setting if you're certain you don't need it.

For more information, see Network Plugin Requirements or the documentation for your specific container runtime.

Forwarding IPv4 and letting iptables see bridged traffic

Execute the below mentioned instructions:

cat <<EOF | sudo tee /etc/modules-load.d/k8s.conf
overlay
br_netfilter
EOF

sudo modprobe overlay
sudo modprobe br_netfilter

# sysctl params required by setup, params persist across reboots
cat <<EOF | sudo tee /etc/sysctl.d/k8s.conf
net.bridge.bridge-nf-call-iptables  = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.ipv4.ip_forward                 = 1
EOF

# Apply sysctl params without reboot
sudo sysctl --system

Verify that the br_netfilter, overlay modules are loaded by running the following commands:

lsmod | grep br_netfilter
lsmod | grep overlay

Verify that the net.bridge.bridge-nf-call-iptables, net.bridge.bridge-nf-call-ip6tables, and net.ipv4.ip_forward system variables are set to 1 in your sysctl config by running the following command:

sysctl net.bridge.bridge-nf-call-iptables net.bridge.bridge-nf-call-ip6tables net.ipv4.ip_forward

cgroup drivers

On Linux, control groups are used to constrain resources that are allocated to processes.

Both the kubelet and the underlying container runtime need to interface with control groups to enforce resource management for pods and containers and set resources such as cpu/memory requests and limits. To interface with control groups, the kubelet and the container runtime need to use a cgroup driver. It's critical that the kubelet and the container runtime use the same cgroup driver and are configured the same.

There are two cgroup drivers available:

• cgroupfs
• systemd

cgroupfs driver

The cgroupfs driver is the default cgroup driver in the kubelet. When the cgroupfs driver is used, the kubelet and the container runtime directly interface with the cgroup filesystem to configure cgroups.

The cgroupfs driver is not recommended when systemd is the init system because systemd expects a single cgroup manager on the system. Additionally, if you use cgroup v2, use the systemd cgroup driver instead of cgroupfs.

systemd cgroup driver

When systemd is chosen as the init system for a Linux distribution, the init process generates and consumes a root control group (cgroup) and acts as a cgroup manager.

systemd has a tight integration with cgroups and allocates a cgroup per systemd unit. As a result, if you use systemd as the init system with the cgroupfs driver, the system gets two different cgroup managers.

Two cgroup managers result in two views of the available and in-use resources in the system. In some cases, nodes that are configured to use cgroupfs for the kubelet and container runtime, but use systemd for the rest of the processes become unstable under resource pressure.

The approach to mitigate this instability is to use systemd as the cgroup driver for the kubelet and the container runtime when systemd is the selected init system.

To set systemd as the cgroup driver, edit the KubeletConfiguration option of cgroupDriver and set it to systemd. For example:

apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration
...
cgroupDriver: systemd

In Kubernetes v1.28, with the KubeletCgroupDriverFromCRI feature gate enabled and a container runtime that supports the RuntimeConfig CRI RPC, the kubelet automatically detects the appropriate cgroup driver from the runtime, and ignores the cgroupDriver setting within the kubelet configuration.

If you configure systemd as the cgroup driver for the kubelet, you must also configure systemd as the cgroup driver for the container runtime. Refer to the documentation for your container runtime for instructions. For example:

• containerd
• CRI-O

Migrating to the systemd driver in kubeadm managed clusters

If you wish to migrate to the systemd cgroup driver in existing kubeadm managed clusters, follow configuring a cgroup driver.

CRI version support

Your container runtime must support at least v1alpha2 of the container runtime interface.

Kubernetes starting v1.26 only works with v1 of the CRI API. Earlier versions default to v1 version, however if a container runtime does not support the v1 API, the kubelet falls back to using the (deprecated) v1alpha2 API instead.

Container runtimes

containerd

This section outlines the necessary steps to use containerd as CRI runtime.

To install containerd on your system, follow the instructions on getting started with containerd. Return to this step once you've created a valid config.toml configuration file.

• Linux
• Windows

On Linux the default CRI socket for containerd is /run/containerd/containerd.sock. On Windows the default CRI endpoint is npipe://./pipe/containerd-containerd.

Configuring the systemd cgroup driver

To use the systemd cgroup driver in /etc/containerd/config.toml with runc, set

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

The systemd cgroup driver is recommended if you use cgroup v2.

If you apply this change, make sure to restart containerd:

sudo systemctl restart containerd

When using kubeadm, manually configure the cgroup driver for kubelet.

In Kubernetes v1.28, you can enable automatic detection of the cgroup driver as an alpha feature. See systemd cgroup driver for more details.

Overriding the sandbox (pause) image

In your containerd config you can overwrite the sandbox image by setting the following config:

[plugins."io.containerd.grpc.v1.cri"]
  sandbox_image = "registry.k8s.io/pause:3.2"

You might need to restart containerd as well once you've updated the config file: systemctl restart containerd.

Please note, that it is a best practice for kubelet to declare the matching pod-infra-container-image. If not configured, kubelet may attempt to garbage collect the pause image. There is ongoing work in containerd to pin the pause image and not require this setting on kubelet any longer.

CRI-O

This section contains the necessary steps to install CRI-O as a container runtime.

To install CRI-O, follow CRI-O Install Instructions.

cgroup driver

CRI-O uses the systemd cgroup driver per default, which is likely to work fine for you. To switch to the cgroupfs cgroup driver, either edit /etc/crio/crio.conf or place a drop-in configuration in /etc/crio/crio.conf.d/02-cgroup-manager.conf, for example:

[crio.runtime]
conmon_cgroup = "pod"
cgroup_manager = "cgroupfs"

You should also note the changed conmon_cgroup, which has to be set to the value pod when using CRI-O with cgroupfs. It is generally necessary to keep the cgroup driver configuration of the kubelet (usually done via kubeadm) and CRI-O in sync.

In Kubernetes v1.28, you can enable automatic detection of the cgroup driver as an alpha feature. See systemd cgroup driver for more details.

For CRI-O, the CRI socket is /var/run/crio/crio.sock by default.

Overriding the sandbox (pause) image

In your CRI-O config you can set the following config value:

[crio.image]
pause_image="registry.k8s.io/pause:3.6"

This config option supports live configuration reload to apply this change: systemctl reload crio or by sending SIGHUP to the crio process.

Docker Engine

1. On each of your nodes, install Docker for your Linux distribution as per Install Docker Engine.

2. Install cri-dockerd, following the instructions in that source code repository.

For cri-dockerd, the CRI socket is /run/cri-dockerd.sock by default.

Mirantis Container Runtime

Mirantis Container Runtime (MCR) is a commercially available container runtime that was formerly known as Docker Enterprise Edition.

You can use Mirantis Container Runtime with Kubernetes using the open source cri-dockerd component, included with MCR.

To learn more about how to install Mirantis Container Runtime, visit MCR Deployment Guide.

Check the systemd unit named cri-docker.socket to find out the path to the CRI socket.

Overriding the sandbox (pause) image

The cri-dockerd adapter accepts a command line argument for specifying which container image to use as the Pod infrastructure container (“pause image”). The command line argument to use is --pod-infra-container-image.

What's next

As well as a container runtime, your cluster will need a working network plugin.