我正在尝试编写一个ansible playbook来设置一些docker容器然后在它们上面运行一个角色:
- hosts: localhost
gather_facts: no
vars:
- docker_test_hosts:
- container_name: 'test_precise'
image_name: 'ubuntu'
image_tag: '12.04'
- container_name: 'test_trusty'
image_name: 'ubuntu'
image_tag: '14.04'
# "Registry returned more than one result for ubuntu:16.04"
#- container_name: 'xenial'
#- image_name: 'ubuntu'
#- image_tag: '16.04'
tasks:
- pip:
name: docker-py
# >= 1.7.0
# using 1.9.0 due to https://github.com/ansible/ansible/issues/17495
version: 1.9.0
state: present
- docker_container:
name: '{{item.container_name}}'
image: '{{item.image_name}}:{{item.image_tag}}'
pull: yes
with_items: '{{docker_test_hosts}}'
- add_host:
name: '{{item.container_name}}'
hostname: '{{item.container_name}}'
ansible_host: '{{item.container_name}}'
ansible_connection: docker
ansible_user: root
groups: docker
with_items: '{{docker_test_hosts}}'
- hosts: docker
tasks:
- debug:
msg: 'hello'
第二场比赛一直失败:
播放[localhost]
[...]
任务[add_host] 已更改:[localhost] => (item = {u' image_tag&#39 ;: u' 12.04',u' image_name':u' ubuntu',u' container_name': U' test_precise'}) 已更改:[localhost] => (item = {u' image_tag&#39 ;: u' 14.04',u' image_name':u' ubuntu',u' container_name': U' test_trusty'})
播放[docker]
任务[设置] 致命:[test_precise]:无法访问! => {"已更改":false," msg":"身份验证或权限失败。在某些情况下,您可能已经能够对远程目录进行身份验证并且没有权限。考虑将ansible.cfg中的远程临时路径更改为以" / tmp \"为根的路径。失败的命令是:(umask 77&& mkdir -p \"
echo $HOME/.ansible/tmp/ansible/tmp-1474479086.86-239783828445202
\"&& echo ansible-tmp-1474479086.86-239783828445202 = \" {{1退出结果1","无法访问":true}[...]
有什么想法吗?
docker 1.11.2 ansible 2.1.1.0 python 2.7.12 Linux Mint 18 Sarah
答案 0 :(得分:3)
首先:你需要让docker容器保持运行,所以使用
- docker_container:
name: '{{item.container_name}}'
image: '{{item.image_name}}:{{item.image_tag}}'
command: tail -f /dev/null
pull: yes
应该导致更改的错误消息:fatal: [test_trusty]: FAILED! => {"changed": false, "failed": true, "module_stderr": "/bin/sh: 1: /usr/bin/python: not found\n", "module_stdout": "", "msg": "MODULE FAILURE", "parsed": false}
这意味着python没有安装在容器中。所以在使用容器之前,需要在其中安装python。您可以通过自定义Dockerfile
并使用创建的泊坞窗图像而不是默认的ubuntu图像来执行此操作