订阅内容

Editor's note: This article covers utilizing the Ansible docker_image module for working with containers. But if you're using a different container toolchain, you may have other options. Be sure to check out our full collection of container articles for the latest and greatest tools for managing Linux containers.

Building container images and placing them into upstream repositories is a common way of making your images available to the outside world. However, you might encounter scenarios where you don't want to upload a container image to a repository. Maybe you don't want to upload your image for security reasons, or maybe your cluster is just sufficiently small (e.g., in your home lab) that you don't want to deal with a repository (whether external or self-hosted).

If you have a use case for avoiding the storage of your container images in a repository, then you're in luck! The Ansible docker_image module makes it easy to build, save, and load your images without ever hitting a repository. This article walks you through some simple playbooks that you can incorporate into your workflow to manage containers with Ansible.

Environment overview

First, let me introduce the basic directory structure that I’ll be using for this tutorial:

$ ls -lah
total 28K
drwxr-xr-x  2 acritelli acritelli 4.0K May  6 11:24 ./
drwxr-xr-x 29 acritelli acritelli 4.0K May  3 08:26 ../
-rw-rw-r--  1 acritelli acritelli  261 May  2 22:23 save.yml
-rw-rw-r--  1 acritelli acritelli  254 May  2 22:20 build.yml
-rw-rw-r--  1 acritelli acritelli   49 May  2 22:01 Dockerfile
-rw-rw-r--  1 acritelli acritelli   54 May  2 22:45 inventory.ini
-rw-rw-r--  1 acritelli acritelli  403 May  2 22:49 load.yml

I have a Dockerfile, an Ansible inventory, and a few simple playbooks for working with the Ansible Docker image module. The hosts in this environment are defined in the inventory.ini file:

$ cat inventory.ini
[build_host]
docker-build.example.com

[docker_hosts]
docker01.example.com
docker02.example.com

There are three hosts that I am working with in this article:

  • docker-build - This server is used to build Docker containers. The container images are then saved off this host.
  • docker01 and docker02 - These hosts are used to run containers. Containers from build01 are uploaded and made available to both of these hosts.

All three hosts in the environment have Docker and the Docker SDK for Python installed. The SDK is a requirement for using the Ansible Docker modules. You can install all of these from the standard repositories via yum:

[root@docker-build ~]# yum install docker python-docker-py

Building a container

The first step in this workflow is building an actual container image. The image that I use in this article is very simple: it just launches a netcat listener on port 8080 and waits for client connections, as you can see from the Dockerfile:

$ cat Dockerfile
FROM alpine:latest
EXPOSE 8080
CMD nc -l -p 8080

A simple Ansible playbook can then be used to build a container image based on this Dockerfile:

$ cat build.yml
---
- hosts: build_host
  gather_facts: no
  tasks:
    - name: create build directory
      file:
        path: /root/demo-dockerfile
        state: directory
        owner: root
        group: root
        mode: '0755'
    - name: copy Dockerfile
      copy:
        src: ./Dockerfile
        dest: /root/demo-dockerfile/Dockerfile
        owner: root
        group: root
        mode: '0644'
    - name: build container image
      docker_image:
        name: democontainer:v1.0
        build:
          path: /root/demo-dockerfile
          source: build
        state: present

This playbook creates a build directory on the "build" server, copies the Dockerfile to this directory, and then builds the container using the docker_image Ansible module. The source parameter set to "build" tells the module to build a container image based on the listed path. In this case, the path points to the directory and Dockerfile that were copied to the build host.

Running this playbook successfully builds the container image:

$ ansible-playbook -i inventory.ini build.yml

PLAY [build_host] ****************************************************************************************************************************************************************************

TASK [create build directory] ****************************************************************************************************************************************************************
changed: [docker-build.example.com]

TASK [copy Dockerfile] ***********************************************************************************************************************************************************************
changed: [docker-build.example.com]

TASK [build container image] *****************************************************************************************************************************************************************
changed: [docker-build.example.com]

PLAY RECAP ***********************************************************************************************************************************************************************************
docker-build.example.com   : ok=3    changed=3    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   

You can log into the build server after the playbook completes to see that the democontainer:v1.0 image has been successfully built and is available to launch a container:

[root@docker-build ~]# docker image ls
REPOSITORY          TAG                 IMAGE ID            CREATED              SIZE
democontainer       v1.0                f2d7d737743e        About a minute ago   5.61 MB
docker.io/alpine    latest              f70734b6a266        13 days ago          5.61 MB

Saving a container image

Container images can be saved to a tarball using the docker save command. Tarballs provide a convenient way to "export" your container images. The Ansible Docker image module also includes support to export an image to a tar file:

$ cat save.yml
---
- hosts: build_host
  gather_facts: no
  tasks:
    - name: archive container image as a tarball
      docker_image:
        name: democontainer:v1.0
        archive_path: /root/democontainer_v1_0.tar
        source: pull
        state: present
    - name: fetch archived image
      fetch:
        src: /root/democontainer_v1_0.tar
        dest: ./democontainer_v1_0.tar
        flat: true

This playbook first archives the image using the docker_image module and then fetches the file from the remote server and places it into the local directory. After successfully running the playbook, you have a tar file in your local directory that contains the contents of the image:

$ ansible-playbook -i inventory.ini save.yml

PLAY [build_host] ****************************************************************************************************************************************************************************

TASK [archive container image as a tarball] **************************************************************************************************************************************************
changed: [docker-build.example.com]

TASK [fetch archived image] ******************************************************************************************************************************************************************
changed: [docker-build.example.com]

PLAY RECAP ***********************************************************************************************************************************************************************************
docker-build.example.com   : ok=2    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0

Notice that the democontainer is now tarred up in the local directory:

$ ls
ansible.cfg  save.yml  build.yml  democontainer_v1_0.tar  Dockerfile  inventory.ini  load.yml

Loading a container image

With the image now downloaded to your local system, you can again use the Ansible Docker image module to upload the tarball to all of your Docker hosts and import it. Once imported, the container image is available to launch containers.

$ cat load.yml
---
- hosts: docker_hosts
  gather_facts: no
  tasks:
    - name: copy tarball to host
      copy:
        src: ./democontainer_v1_0.tar
        dest: /root/democontainer_v1_0.tar

    - name: load container from tarball
      docker_image:
        name: democontainer:v1.0
        load_path: /root/democontainer_v1_0.tar
        state: present
        source: load

This playbook copies the tarball created in the save.yml playbook to all of the Docker hosts. It then loads the container image to make it available to future docker run commands.

$ ansible-playbook -i inventory.ini load.yml

PLAY [docker_hosts] **************************************************************************************************************************************************************************

TASK [copy tarball to host] ******************************************************************************************************************************************************************
changed: [docker01.example.com]
changed: [docker02.example.com]

TASK [load container from tarball] ***********************************************************************************************************************************************************
changed: [docker01.example.com]
changed: [docker02.example.com]

PLAY RECAP ***********************************************************************************************************************************************************************************
docker01.example.com       : ok=2    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   
docker02.example.com       : ok=2    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0

Once the playbook has successfully run, the container image can be seen when listing out the available images on any of the Docker hosts:

[root@docker01 ~]# docker image ls
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
democontainer       v1.0                f2d7d737743e        25 hours ago        5.61 MB
[root@docker02 ~]# docker image ls
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
democontainer       v1.0                f2d7d737743e        25 hours ago        5.61 MB

You can then launch a container on one of the hosts and see that it works.

Launch the container:

[root@docker01 ~]# docker run --rm -it -p 8080:8080 democontainer:v1.0

The below command is executed in a separate session:

[root@docker01 ~]# telnet localhost 8080
Trying ::1...
Connected to localhost.
Escape character is '^]'.
hello!

You can see that the output is echoed back within the container via netcat:

[root@docker01 ~]# docker run --rm -it -p 8080:8080 democontainer:v1.0
hello!

If you want to avoid the need to copy the container image to your local host and then copy it back to the Docker hosts, you can check out the Ansible synchronize module. While out of scope for this article, you can gain some efficiencies with this thin wrapper around rsync.

Wrapping up

This article walked you through the use of the Ansible Docker image module, which provides a way to manage container images on remote hosts. You learned how to build, save, and load an image through simple Ansible playbooks. While using a container repository is a standard way to make an image available across multiple hosts, the use of this Ansible module can also provide a simple mechanism for sharing your images in a small environment.

[ Need more on Ansible? Take a free technical overview course from Red Hat. Ansible Essentials: Simplicity in Automation Technical Overview. ]


关于作者

Anthony Critelli is a Linux systems engineer with interests in automation, containerization, tracing, and performance. He started his professional career as a network engineer and eventually made the switch to the Linux systems side of IT. He holds a B.S. and an M.S. from the Rochester Institute of Technology.

Read full bio
UI_Icon-Red_Hat-Close-A-Black-RGB

按频道浏览

automation icon

自动化

有关技术、团队和环境 IT 自动化的最新信息

AI icon

人工智能

平台更新使客户可以在任何地方运行人工智能工作负载

open hybrid cloud icon

开放混合云

了解我们如何利用混合云构建更灵活的未来

security icon

安全防护

有关我们如何跨环境和技术减少风险的最新信息

edge icon

边缘计算

简化边缘运维的平台更新

Infrastructure icon

基础架构

全球领先企业 Linux 平台的最新动态

application development icon

应用领域

我们针对最严峻的应用挑战的解决方案

Original series icon

原创节目

关于企业技术领域的创客和领导者们有趣的故事