Docker

Table of Contents

1. Description
2. Setup
   • Proxy on Windows
   • Validating and unit testing the module
3. Usage - Configuration options and additional functionality
   • Images
   • Containers
   • Networks
   • Volumes
   • Compose
   • Machine
   • Swarm mode
   • Tasks
   • Docker services
   • Private registries
   • Exec
   • Plugins
4. Reference - An under-the-hood peek at what the module is doing and how
5. Limitations - OS compatibility, etc.
6. Development - Guide for contributing to the module
7. Acceptance

Description

The Puppet docker module installs, configures, and manages Docker from the Docker repository. It supports the latest Docker CE (Community Edition) for Linux based distributions and Docker EE(Enterprise Edition) for Windows and Linux as well as legacy releases.

Due to the new naming convention for Docker packages, this module prefaces any params that refer to the release with _ce or _engine. Examples of these are documented in this README.

Setup

To create the Docker hosted repository and install the Docker package, add a single class to the manifest file:

include 'docker'

To configure package sources independently and disable automatically including sources, add the following code to the manifest file:

class { 'docker':
  use_upstream_package_source => false,
}

The latest Docker repositories are now the default repositories for version 17.06 and above. If you are using an older version, the repositories are still configured based on the version number passed into the module.

To ensure the module configures the latest repositories, add the following code to the manifest file:

class { 'docker':
  version => '17.09.0~ce-0~debian',
}

Using a version prior to 17.06, configures and installs from the old repositories:

class { 'docker':
  version => '1.12.0-0~wheezy',
}

Docker provides a enterprise addition of the Docker Engine, called Docker EE. To install Docker EE on Debian systems, add the following code to the manifest file:

class { 'docker':
  docker_ee                 => true,
  docker_ee_source_location => 'https://<docker_ee_repo_url>',
  docker_ee_key_source      => 'https://<docker_ee_key_source_url>',
  docker_ee_key_id          => '<key id>',
}

To install Docker EE on RHEL/CentOS:

class { 'docker':
  docker_ee                 => true,
  docker_ee_source_location => 'https://<docker_ee_repo_url>',
  docker_ee_key_source      => 'https://<docker_ee_key_source_url>',
}

For CentOS distributions, the docker module requires packages from the extras repository, which is enabled by default on CentOS. For more information, see the official CentOS documentation and the official Docker documentation.

For Red Hat Enterprise Linux (RHEL) based distributions, the docker module uses the upstream repositories. To continue using the legacy distribution packages in the CentOS extras repository, add the following code to the manifest file:

class { 'docker':
  use_upstream_package_source => false,
  service_overrides_template  => false,
  docker_ce_package_name      => 'docker',
}

To use the CE packages, add the following code to the manifest file:

class { 'docker':
  use_upstream_package_source => false,
  repo_opt                    => '',
}

By default, the Docker daemon binds to a unix socket at /var/run/docker.sock. To change this parameter and update the binding parameter to a tcp socket, add the following code to the manifest file:

class { 'docker':
  tcp_bind    => ['tcp://127.0.0.1:2375'],
  socket_bind => 'unix:///var/run/docker.sock',
  ip_forward  => true,
  iptables    => true,
  ip_masq     => true,
  bip         => '192.168.1.1/24',
  fixed_cidr  => '192.168.1.144/28',
}

For more information about the configuration options for the default docker bridge, see the Docker documentation.

The default group ownership of the Unix control socket differs based on OS. For example, on RHEL using docker-ce packages >=18.09.1, the socket file used by /usr/lib/systemd/system/docker.socket is owned by the docker group. To override this value in /etc/sysconfig/docker and docker.socket (e.g. to use the 'root' group):

class {'docker':
  socket_group    => 'root',
  socket_override => true,
}

The socket_group parameter also takes a boolean for legacy cases where setting -G in /etc/sysconfig/docker is not desired:

docker::socket_group: false

To add another service to the After= line in the [Unit] section of the systemd /etc/systemd/system/service-overrides.conf file, use the service_after_override parameter:

docker::service_after_override: containerd.service

When setting up TLS, upload the related files (CA certificate, server certificate, and key) and include their paths in the manifest file:

class { 'docker':
  tcp_bind   => ['tcp://0.0.0.0:2376'],
  tls_enable => true,
  tls_cacert => '/etc/docker/tls/ca.pem',
  tls_cert   => '/etc/docker/tls/cert.pem',
  tls_key    => '/etc/docker/tls/key.pem',
}

To specify which Docker rpm package to install, add the following code to the manifest file:

class { 'docker':
  manage_package              => true,
  use_upstream_package_source => false,
  package_engine_name         => 'docker-engine'
  package_source_location     => 'https://get.docker.com/rpm/1.7.0/centos-6/RPMS/x86_64/docker-engine-1.7.0-1.el6.x86_64.rpm',
  prerequired_packages        => [ 'glibc.i686', 'glibc.x86_64', 'sqlite.i686', 'sqlite.x86_64', 'device-mapper', 'device-mapper-libs', 'device-mapper-event-libs', 'device-mapper-event' ]
}

To track the latest version of Docker, add the following code to the manifest file:

class { 'docker':
  version => latest,
}

To install docker from a test or edge channel, add the following code to the manifest file:

class { 'docker':
  docker_ce_channel => 'test'
}

To allocate a DNS server to the Docker daemon, add the following code to the manifest file:

class { 'docker':
  dns => '8.8.8.8',
}

To add users to the Docker group, add the following array to the manifest file:

class { 'docker':
  docker_users => ['user1', 'user2'],
}

To add daemon labels, add the following array to the manifest file:

class { 'docker':
  labels => ['storage=ssd','stage=production'],
}

To pass additional parameters to the daemon, add extra_parameters to the manifest file:

class { 'docker':
  extra_parameters => ['--experimental=true', '--metrics-addr=localhost:9323'],

To uninstall docker, add the following to the manifest file:

class { 'docker':
  ensure => absent
}

Only Docker EE is supported on Windows. To install docker on Windows 2016 and above, the docker_ee parameter must be specified:

class { 'docker':
  docker_ee => true
}

If the curl package is being managed elsewhere and the curl ensure in this module is conflicting, it can be disabled by setting the following parameter globally or in compose / machine resources:

class { 'docker':
  curl_ensure => false
}

Proxy on Windows

To use docker through a proxy on Windows, a System Environment Variable HTTP_PROXY/HTTPS_PROXY must be set. See Docker Engine on Windows This can be done using a different puppet module, such as the puppet-windows_env module. After setting the variable, the docker service must be restarted.

windows_env { 'HTTP_PROXY'
  value  => 'http://1.2.3.4:80',
  notify => Service['docker'],
}
windows_env { 'HTTPS_PROXY'
  value  => 'http://1.2.3.4:80',
  notify => Service['docker'],
}
service { 'docker'
  ensure => 'running',
}

Validating and unit testing the module

This module is compliant with the Puppet Development Kit (PDK), which provides tools to help run unit tests on the module and validate the modules' metadata, syntax, and style.

To run all validations against this module, run the following command:

pdk validate

To change validation behavior, add options flags to the command. For a complete list of command options and usage information, see the PDK command reference.

To unit test the module, run the following command:

pdk test unit

To change unit test behavior, add option flags to the command. For a complete list of command options and usage information, see the PDK command reference.

Usage

Images

Each image requires a unique name; otherwise, the installation fails when a duplicate name is detected.

To install a Docker image, add the docker::image defined type to the manifest file:

docker::image { 'base': }

The code above is equivalent to running the docker pull base command. However, it removes the default five-minute execution timeout.

To include an optional parameter for installing image tags that is the equivalent to running docker pull -t="precise" ubuntu, add the following code to the manifest file:

docker::image { 'ubuntu':
  image_tag => 'precise'
}

Including the docker_file parameter is equivalent to running the docker build -t ubuntu - < /tmp/Dockerfile command. To add or build an image from a dockerfile that includes the docker_file parameter, add the following code to the manifest file:

docker::image { 'ubuntu':
  docker_file => '/tmp/Dockerfile'
}

Including the docker_dir parameter is equivalent to running the docker build -t ubuntu /tmp/ubuntu_image command. To add or build an image from a dockerfile that includes the docker_dir parameter, add the following code to the manifest file:

docker::image { 'ubuntu':
  docker_dir => '/tmp/ubuntu_image'
}

To rebuild an image, subscribe to external events such as Dockerfile changes by adding the following code to the manifest file:

docker::image { 'ubuntu':
  docker_file => '/tmp/Dockerfile',
  subscribe   => File['/tmp/Dockerfile'],
}

file { '/tmp/Dockerfile':
  ensure => file,
  source => 'puppet:///modules/someModule/Dockerfile',
}

To remove an image, add the following code to the manifest file:

docker::image { 'base':
  ensure => 'absent'
}

docker::image { 'ubuntu':
  ensure    => 'absent',
  image_tag => 'precise'
}

To configure the docker::images class when using Hiera, add the following code to the manifest file:

---
  classes:
    - docker::images

docker::images::images:
  ubuntu:
    image_tag: 'precise'

Containers

To launch containers, add the following code to the manifest file:

docker::run { 'helloworld':
  image   => 'base',
  command => '/bin/sh -c "while true; do echo hello world; sleep 1; done"',
}

This is equivalent to running the docker run -d base /bin/sh -c "while true; do echo hello world; sleep 1; done" command to launch a Docker container managed by the local init system.

run includes a number of optional parameters:

docker::run { 'helloworld':
  image            => 'base',
  detach           => true,
  service_prefix   => 'docker-',
  command          => '/bin/sh -c "while true; do echo hello world; sleep 1; done"',
  ports            => ['4444', '4555'],
  expose           => ['4666', '4777'],
  links            => ['mysql:db'],
  net              => ['my-user-def-net','my-user-def-net-2'],
  disable_network  => false,
  volumes          => ['/var/lib/couchdb', '/var/log'],
  volumes_from     => '6446ea52fbc9',
  memory_limit     => '10m', # (format: '<number><unit>', where unit = b, k, m or g)
  cpuset           => ['0', '3'],
  username         => 'example',
  hostname         => 'example.com',
  env              => ['FOO=BAR', 'FOO2=BAR2'],
  env_file         => ['/etc/foo', '/etc/bar'],
  labels           => ['com.example.foo="true"', 'com.example.bar="false"'],
  dns              => ['8.8.8.8', '8.8.4.4'],
  restart_service  => true,
  privileged       => false,
  pull_on_start    => false,
  before_stop      => 'echo "So Long, and Thanks for All the Fish"',
  before_start     => 'echo "Run this on the host before starting the Docker container"',
  after_stop       => 'echo "container has stopped"',
  after_start      => 'echo "container has started"',
  after            => [ 'container_b', 'mysql' ],
  depends          => [ 'container_a', 'postgres' ],
  stop_wait_time   => 0,
  read_only        => false,
  extra_parameters => [ '--restart=always' ],
}

You can specify the ports, expose, env, dns, and volumes values with a single string or an array.

To pull the image before it starts, specify the pull_on_start parameter.

Use the detach param to run a container without the -a flag. This is only required on systems without systemd. This default is set in the params.pp based on the OS. Only override if you understand the consequences and have a specific use case.

To execute a command before the container starts or stops, specify the before_start or before_stop parameters, respectively. Similarly, you can set the after_start or after_stop parameters to run a command after the container starts or stops.

Adding the container name to the after parameter to specify which containers start first affects the generation of the init.d/systemd script.

Add container dependencies to the depends parameter. The container starts before this container and stops before the dependent container. This affects the generation of the init.d/systemd script. Use the depend_services parameter to specify dependencies for general services, which are not Docker related, that start before this container.

The extra_parameters parameter, which contains an array of command line arguments to pass to the docker run command, is useful for adding additional or experimental options that the docker module currently does not support.

By default, automatic restarting of the service on failure is enabled by the service file for systemd based systems.

It's recommended that an image tag is used at all times with the docker::run define type. If not, the latest image is used whether it's in a remote registry or installed on the server already by the docker::image define type.

NOTE: As of v3.0.0, if the latest tag is used the image will be the latest at the time the of the initial puppet run. Any subsequent puppet runs will always reference the latest local image. Therefore, it's recommended that an alternative tag be used, or the image be removed before pulling latest again.

To use an image tag, add the following code to the manifest file:

docker::run { 'helloworld':
  image   => 'ubuntu:precise',
  command => '/bin/sh -c "while true; do echo hello world; sleep 1; done"',
}

By default, when the service stops or starts, the generated init scripts remove the container, but not the associated volumes. To change this behaviour, add the following code to the manifest file:

docker::run { 'helloworld':
  remove_container_on_start => true,
  remove_volume_on_start    => false,
  remove_container_on_stop  => true,
  remove_volume_on_stop     => false,
}

If using Hiera, you can configure the docker::run_instance class:

---
  classes:
    - docker::run_instance

  docker::run_instance::instance:
    helloworld:
      image: 'ubuntu:precise'
      command: '/bin/sh -c "while true; do echo hello world; sleep 1; done"'

To remove a running container, add the following code to the manifest file. This also removes the systemd service file associated with the container.

docker::run { 'helloworld':
  ensure => absent,
}

To enable the restart of an unhealthy container, add the following code to the manifest file. To set the health check interval time, set the optional health_check_interval parameter. The default health check interval is 30 seconds.

docker::run { 'helloworld':
  image                 => 'base',
  command               => 'command',
  health_check_cmd      => '<command_to_execute_to_check_your_containers_health>',
  restart_on_unhealthy  => true,
  health_check_interval => '<time between running docker healthcheck>',

To run command on Windows 2016 requires the restart parameter to be set:

docker::run { 'helloworld':
  image   => 'microsoft/nanoserver',
  command => 'ping 127.0.0.1 -t',
  restart => 'always'

Networks

Docker 1.9.x supports networks. To expose the docker_network type that is used to manage networks, add the following code to the manifest file:

docker_network { 'my-net':
  ensure   => present,
  driver   => 'overlay',
  subnet   => '192.168.1.0/24',
  gateway  => '192.168.1.1',
  ip_range => '192.168.1.4/32',
}

The name value and the ensure parameter are required. If you do not include the driver value, the default bridge is used. The Docker daemon must be configured for some networks, and configuring the cluster store for the overlay network would be an example.

To configure the cluster store, update the docker class in the manifest file:

extra_parameters => '--cluster-store=<backend>://172.17.8.101:<port> --cluster-advertise=<interface>:2376'

If using Hiera, configure the docker::networks class in the manifest file:

---
  classes:
    - docker::networks

docker::networks::networks:
  local-docker:
    ensure: 'present'
    subnet: '192.168.1.0/24'
    gateway: '192.168.1.1'

A defined network can be used on a docker::run resource with the net parameter.

Windows

On Windows, only one NAT network is supported. To support multiple networks, Windows Server 2016 with KB4015217 is required. See Windows Container Network Drivers and Windows Container Networking.

The Docker daemon will create a default NAT network on the first start unless specified otherwise. To disable the network creation, use the parameter bridge => 'none' when installing docker.

Volumes

Docker 1.9.x added support for volumes. These are NOT to be confused with the legacy volumes, now known as bind mounts. To expose the docker_volume type, which is used to manage volumes, add the following code to the manifest file:

docker_volume { 'my-volume':
  ensure => present,
}

You can pass additional mount options to the local driver. For mounting an NFS export, use:

docker_volume { 'nfs-volume':
  ensure  => present,
  driver  => 'local',
  options => ['type=nfs','o=addr=%{custom_manager},rw','device=:/srv/blueocean']
}

The name value and the ensure parameter are required. If you do not include the driver value, the default local is used.

If using Hiera, configure the docker::volumes class in the manifest file:

---
  classes:
    - docker::volumes

docker::volumes::volumes:
  blueocean:
    ensure: present
    driver: local
    options:
      - ['type=nfs','o=addr=%{custom_manager},rw','device=:/srv/blueocean']

Available parameters for options depend on the used volume driver. For details, see Using volumes from the Docker manual.

Some of the key advantages for using volumes over bind mounts are:

• Easier to back up or migrate rather than bind mounts (legacy volumes).
• Managed with Docker CLI or API (Puppet type uses the CLI commands).
• Works on Windows and Linux.
• Easily shared between containers.
• Allows for store volumes on remote hosts or cloud providers.
• Encrypt contents of volumes.
• Add other functionality
• New volume's contents can be pre-populated by a container.

When using the volumes array with docker::run, the command on the backend will know if it needs to use bind mounts or volumes based on the data passed to the -v option.

Running docker::run with native volumes:

docker::run { 'helloworld':
  image   => 'ubuntu:precise',
  command => '/bin/sh -c "while true; do echo hello world; sleep 1; done"',
  volumes => ['my-volume:/var/log'],
}

Compose

Docker Compose describes a set of containers in YAML format and runs a command to build and run those containers. Included in the docker module is the docker_compose type. This enables Puppet to run Compose and remediate any issues to ensure reality matches the model in your Compose file.

Before you use the docker_compose type, you must install the Docker Compose utility.

To install Docker Compose, add the following code to the manifest file:

class {'docker::compose':
  ensure  => present,
  version => '1.9.0',
}

Set the version parameter to any version you need to install.

This is an example of a Compose file:

compose_test:
  image: ubuntu:14.04
  command: /bin/sh -c "while true; do echo hello world; sleep 1; done"

Specify the file resource to add a Compose file to the machine you have Puppet running on. To define a docker_compose resource pointing to the Compose file, add the following code to the manifest file:

docker_compose { 'test':
  compose_files => ['/tmp/docker-compose.yml'],
  ensure        => present,
}

Puppet automatically runs Compose because the relevant Compose services aren't running. If required, include additional options such as enabling experimental features and scaling rules.

Additionally, the TMPDIR environment variable can optionally be set when docker_compose runs if you want Puppet to manage the environment variable within the scope of the resource. This is effective when noexec is set on the default /tmp dir, however you must ensure that the target directory exists as the resource will not create it.

In the example below, Puppet runs Compose when the number of containers specified for a service doesn't match the scale values. The optional tmpdir parameter is also specified.

docker_compose { 'test':
  compose_files => ['/tmp/docker-compose.yml'],
  ensure        => present,
  scale         => {
    'compose_test' => 2,
  },
  tmpdir        => '/usr/local/share/tmp_docker',
  options       => ['--x-networking']
}

Give options to the docker-compose up command, such as --remove-orphans, by using the up_args option.

To supply multiple overide compose files add the following to the manifest file:

docker_compose {'test':
  compose_files => ['server-docker-compose.yml', 'override-compose.yml'],
}

Please note you should supply your server docker-compose file as the first element in the array. As per docker, multi compose file support compose files are merged in the order they are specified in the array.

If you are using a v3.2 compose file or above on a Docker Swarm cluster, use the docker::stack class. Include the file resource before you run the stack command.

NOTE: this define will be deprecated in a future release in favor of the docker stack type

To deploy the stack, add the following code to the manifest file:

 docker::stack { 'yourapp':
   ensure        => present,
   stack_name    => 'yourapp',
   compose_files => ['/tmp/docker-compose.yaml'],
   require       => [Class['docker'], File['/tmp/docker-compose.yaml']],
}

To remove the stack, set ensure => absent.

If you are using a v3.2 compose file or above on a Docker Swarm cluster, include the docker::stack class. Similar to using older versions of Docker, compose the file resource before running the stack command.

To deploy the stack, add the following code to the manifest file.

docker::stack { 'yourapp':
  ensure             => present,
  stack_name         => 'yourapp',
  compose_files      => ['/tmp/docker-compose.yaml'],
  with_registry_auth => true,
  require            => [Class['docker'], File['/tmp/docker-compose.yaml']],
}

To use the equivalent type and provider, use the following in your manifest file. For more information on specific parameters see the docker_stack type documentation.

docker_stack { 'test':
  compose_files => ['/tmp/docker-compose.yml'],
  ensure        => present,
  up_args       => '--with-registry-auth',
}

To remove the stack, set ensure => absent.

Machine

You can use Docker Machine to install Docker Engine on virtual hosts and manage the hosts with docker-machine commands. You can also use Machine to create Docker hosts on your local Mac or Windows box, on your company network, in your data center, or on cloud providers like Azure, AWS, or Digital Ocean.

For more information on machines, see the Docker Machines documentation.

This module only installs the Docker Machine utility.

To install Docker Machine, add the following code to the manifest file:

class {'docker::machine':
  ensure  => present,
  version => '1.16.1',
}

Set the version parameter to any version you need to install.

Swarm mode

To natively manage a cluster of Docker Engines known as a swarm, Docker Engine 1.12 includes a swarm mode.

To cluster your Docker engines, use one of the following Puppet resources:

• Swarm manager
• Swarm worker

Windows

To configure swarm, Windows Server 2016 requires KB4015217 and the following firewall ports to be open on all nodes:

• TCP port 2377 for cluster management communications
• TCP and UDP port 7946 for communication among nodes
• UDP port 4789 for overlay network traffic

Swarm manager

To configure the swarm manager, add the following code to the manifest file:

docker::swarm {'cluster_manager':
  init           => true,
  advertise_addr => '192.168.1.1',
  listen_addr    => '192.168.1.1',
}

For a multihomed server and to enable cluster communications between the node, include the advertise_addr and listen_addr parameters.

Swarm worker

To configure the swarm worker, add the following code to the manifest file:

docker::swarm {'cluster_worker':
  join           => true,
  advertise_addr => '192.168.1.2',
  listen_addr    => '192.168.1.2',
  manager_ip     => '192.168.1.1',
  token          => 'your_join_token'
}

To configure a worker node or a second manager, include the swarm manager IP address in the manager_ip parameter. To define the role of the node in the cluster, provide the token parameter. When creating an additional swarm manager and a worker node, separate tokens are required. These tokens (i.e. docker_worker_join_token and docker_manager_join_token) can be retrieved from Facter.

To remove a node from a cluster, add the following code to the manifest file:

docker::swarm {'cluster_worker':
  ensure => absent
}

Tasks

The docker module has an example task that allows a user to initialize, join and leave a swarm.

bolt task run docker::swarm_init listen_addr=172.17.10.101 adverstise_addr=172.17.10.101 ---nodes swarm-master --user <user> --password <password> --modulepath <module_path>

docker swarm init --advertise-addr=172.17.10.101 --listen-addr=172.17.10.101
Swarm initialized: current node (w8syk0g286vd7d9kwzt7jl44z) is now a manager.

To add a worker to this swarm, run the following command:

    docker swarm join --token SWMTKN-1-317gw63odq6w1foaw0xkibzqy34lga55aa5nbjlqekcrhg8utl-08vrg0913zken8h9vfo4t6k0t 172.17.10.101:2377

To add a manager to this swarm, run docker swarm join-token manager and follow the instructions.

Ran on 1 node in 4.04 seconds

bolt task run docker::swarm_token node_role=worker ---nodes swarm-master --user <user> --password <password> --modulepath <module_path>

SWMTKN-1-317gw63odq6w1foaw0xkibzqy34lga55aa5nbjlqekcrhg8utl-08vrg0913zken8h9vfo4t6k0t

Ran on 1 node in 4.02 seconds

bolt task run docker::swarm_join listen_addr=172.17.10.102 adverstise_addr=172.17.10.102 token=<swarm_token> manager_ip=172.17.10.101:2377 --nodes swarm-02 --user root --password puppet --modulepath /tmp/modules

This node joined a swarm as a worker.

Ran on 1 node in 4.68 seconds

bolt task run docker::swarm_leave --nodes swarm-02 --user root --password puppet --modulepath --modulepath <module_path>

Node left the swarm.

Ran on 1 node in 6.16 seconds

Docker services

Docker services create distributed applications across multiple swarm nodes. Each Docker service replicates a set of containers across the swarm.

To create a Docker service, add the following code to the manifest file:

docker::services {'redis':
    create       => true,
    service_name => 'redis',
    image        => 'redis:latest',
    publish      => '6379:639',
    replicas     => '5',
    mounts       => ['type=bind,source=/etc/my-redis.conf,target=/etc/redis/redis.conf,readonly'],
    extra_params => ['--update-delay 1m', '--restart-window 30s'],
    command      => ['redis-server', '--appendonly', 'yes'],
  }

To base the service off an image, include the image parameter and set the publish parameter to expose the service port (use an array to specify multiple published ports). To set the number of containers running in the service, include the replicas parameter. To attach one or multiple filesystems to the service, use the mounts parameter. For information regarding the extra_params parameter, see docker service create --help. The command parameter can either be specified as an array or a string.

To update the service, add the following code to the manifest file:

docker::services {'redis_update':
  create       => false,
  update       => true,
  service_name => 'redis',
  replicas     => '3',
}

To update a service without creating a new one, include the the update => true and create => false parameters.

To scale a service, add the following code to the manifest file:

docker::services {'redis_scale':
  create       => false,
  scale        => true,
  service_name => 'redis',
  replicas     => '10',
}

To scale the service without creating a new one, provide the scale => true parameter and the create => false parameters. In the example above, the service is scaled to 10.

To remove a service, add the following code to the manifest file:

docker::services {'redis':
  create       => false,
  ensure       => 'absent',
  service_name => 'redis',
}

To remove the service from a swarm, include the ensure => absent parameter and the service_name parameter.

Private registries

When a server is not specified, images are pushed and pulled from index.docker.io. To qualify your image name, create a private repository without authentication.

To configure authentication for a private registry, add the following code to the manifest file, depending on what version of Docker you are running. If you are using Docker V1.10 or earlier, specify the docker version in the manifest file:

docker::registry { 'example.docker.io:5000':
  username => 'user',
  password => 'secret',
  email    => '[email protected]',
  version  => '<docker_version>'
}

To pull images from the docker store, use the following as the registry definition with your docker hub credentials.

  docker::registry {'https://index.docker.io/v1/':
    username => 'username',
    password => 'password',
  }

If using hiera, configure the docker::registry_auth class:

docker::registry_auth::registries:
  'example.docker.io:5000':
    username: 'user1'
    password: 'secret'
    email: '[email protected]'
    version: '<docker_version>'

If using Docker V1.11 or later, the docker login email flag has been deprecated. See the docker_change_log.

Add the following code to the manifest file:

docker::registry { 'example.docker.io:5000':
  username => 'user',
  password => 'secret',
}

If using hiera, configure the 'docker::registry_auth' class:

docker::registry_auth::registries:
  'example.docker.io:5000':
    username: 'user1'
    password: 'secret'

To log out of a registry, add the following code to the manifest file:

docker::registry { 'example.docker.io:5000':
  ensure => 'absent',
}

To set a preferred registry mirror, add the following code to the manifest file:

class { 'docker':
  registry_mirror => 'http://testmirror.io'
}

Exec

Within the context of a running container, the docker module supports arbitrary commands:

docker::exec { 'cron_allow_root':
  detach      => true,
  container   => 'mycontainer',
  command     => '/bin/echo root >> /usr/lib/cron/cron.allow',
  onlyif      => 'running',
  tty         => true,
  env         => ['FOO=BAR', 'FOO2=BAR2'],
  unless      => 'grep root /usr/lib/cron/cron.allow 2>/dev/null',
  refreshonly => true,
}

Plugin

The module supports the installation of Docker plugins:

docker::plugin {'foo/fooplugin:latest':
  settings => ['VAR1=test','VAR2=value']
}

To disable an active plugin:

docker::plugin {'foo/fooplugin:latest':
  enabled => false,
}

To remove an active plugin:

docker::plugin {'foo/fooplugin:latest'
  ensure => 'absent',
  force_remove => true,
}

Reference

For information on classes, types, and functions, see the REFERENCE.md.

Limitations

This module supports:

• Centos 7.0
• Debian 8.0
• Debian 9.0
• Debian 10
• RedHat 7.0 - limited support available
• Ubuntu 18.04
• Ubuntu 20.04
• Ubuntu 22.04
• Windows Server 2016 (Docker Enterprise Edition only)

On RedHat 7 the default docker package installs docker server version 1.13.1. The default docker.service uses the docker-storage-service in this version and creates /etc/sysconfig/docker-storage based on the container-storage-setup configuration and /etc/sysconfig/docker-storage-setup file. As the puppetlabs-docker module manages both the docker-storage and docker-storage-setup files it causes a conflict with the container-storage-setup forcing a docker service restart, therefore a workaround was included in the service manifest that disables the service restart on storage configuration changes for this version of docker on RedHat 7. As a side effect of these changes, storage configuration changes with this docker version on RedHat 7 are not picked up by default by the docker.service.

Development

If you would like to contribute to this module, see the guidelines in CONTRIBUTING.MD.

Acceptance

Acceptance tests for this module leverage puppet_litmus. To run the acceptance tests follow the instructions here.