Creating OCI bundles

Since containerd consumes the OCI bundle format containers and configuration will have to be created on the machine that containerd is running on. The easiest way to do this is to download an image with docker and export it.

Setup

First thing we need to do to create a bundle is setup the initial directory structure. Create a directory with a unique name. In this example we will create a redis container. We will create this container in a /containers directory.

mkdir redis

Inside the redis directory create another directory named rootfs

mkdir redis/rootfs

Root Filesystem

Now we need to populate the rootfs directory with the filesystem of a redis container. To do this we need to pull the redis image with docker and export its contents to the rootfs directory.

docker pull redis

# create the container with a temp name so that we can export it
docker create --name tempredis redis

# export it into the rootfs directory
docker export tempredis | tar -C redis/rootfs -xf -

# remove the container now that we have exported
docker rm tempredis

Now that we have the root filesystem populated we need to create the configs for the container.

Configs

An easy way to get temp configs for the container bundle is to use the runc cli tool from the runc repository.

You need to cd into the redis directory and run the runc spec command. After doing this you should have a file config.json created. The directory structure should look like this:

/containers/redis
├── config.json
└── rootfs/

Edits

We need to edit the config to add redis-server as the application to launch inside the container, and remove the network namespace so that you can connect to the redis server on your system. The resulting config.json should look like this:

{
	"ociVersion": "0.4.0",
	"platform": {
		"os": "linux",
		"arch": "amd64"
	},
	"process": {
		"terminal": true,
		"user": {},
		"args": [
			"redis-server", "--bind", "0.0.0.0"
		],
		"env": [
			"PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
			"TERM=xterm"
		],
		"cwd": "/",
		"capabilities": [
			"CAP_AUDIT_WRITE",
			"CAP_KILL",
			"CAP_NET_BIND_SERVICE"
		],
		"rlimits": [
			{
				"type": "RLIMIT_NOFILE",
				"hard": 1024,
				"soft": 1024
			}
		],
		"noNewPrivileges": true
	},
	"root": {
		"path": "rootfs",
		"readonly": true
	},
	"hostname": "runc",
	"mounts": [
		{
			"destination": "/proc",
			"type": "proc",
			"source": "proc"
		},
		{
			"destination": "/dev",
			"type": "tmpfs",
			"source": "tmpfs",
			"options": [
				"nosuid",
				"strictatime",
				"mode=755",
				"size=65536k"
			]
		},
		{
			"destination": "/dev/pts",
			"type": "devpts",
			"source": "devpts",
			"options": [
				"nosuid",
				"noexec",
				"newinstance",
				"ptmxmode=0666",
				"mode=0620",
				"gid=5"
			]
		},
		{
			"destination": "/dev/shm",
			"type": "tmpfs",
			"source": "shm",
			"options": [
				"nosuid",
				"noexec",
				"nodev",
				"mode=1777",
				"size=65536k"
			]
		},
		{
			"destination": "/dev/mqueue",
			"type": "mqueue",
			"source": "mqueue",
			"options": [
				"nosuid",
				"noexec",
				"nodev"
			]
		},
		{
			"destination": "/sys",
			"type": "sysfs",
			"source": "sysfs",
			"options": [
				"nosuid",
				"noexec",
				"nodev",
				"ro"
			]
		},
		{
			"destination": "/sys/fs/cgroup",
			"type": "cgroup",
			"source": "cgroup",
			"options": [
				"nosuid",
				"noexec",
				"nodev",
				"relatime",
				"ro"
			]
		}
	],
	"hooks": {},
	"linux": {
		"resources": {
			"devices": [
				{
					"allow": false,
					"access": "rwm"
				}
			]
		},
		"namespaces": [
			{
				"type": "pid"
			},
			{
				"type": "ipc"
			},
			{
				"type": "uts"
			},
			{
				"type": "mount"
			}
		],
		"devices": null
	}
}

This is what you need to do to make a OCI compliant bundle for containerd to start.