Question

我有500多个目标服务器，它们具有兼容的python version 2.7 安装在如下所示的不同位置。

/ usr / bin / python /opt/apps/xerto/tools/ansible/bin/python2.7

/opt/atlassian/tools/ansible/bin/python2.7

/opt/wsp/apps/ansible/python2.7/bin/python2.7

/opt/wsp/apps/tools/ansible/python2.7/bin/python2.7

/opt/docs/python/bin/python2.7

/home/admin/ansible/bin/python2.7

/opt/operations/Python-2.7.8/bin/python

/opt/ora/python/bin/python2.7

/opt/tomcat/tools/ansible/bin/python2.7

每次我都必须在ansible主机文件中为ansible_python_interpreter设置上述python路径之一，具体取决于我的ansible将连接的目标服务器。

我的目的是在目标服务器上进行很少或根本不做任何更改，而是在敏感的一端解决此问题。

是否有一种聪明的方法来弄清楚所需的python在哪里？

请提出建议。

Ansible version: 2.7.1

Answer 1

是否有一种聪明的方法来弄清楚所需的python在哪里？

- hosts: all
  # since we have no working python yet, don't bother
  gather_facts: no
  vars:
     possible_pythons:
     - /usr/bin/python
     - /opt/apps/xerto/tools/ansible/bin/python2.7
     - /opt/atlassian/tools/ansible/bin/python2.7
     - /opt/wsp/apps/ansible/python2.7/bin/python2.7
  tasks:
  - raw: if [ -x "{{ item }}" ]; then echo "{{ item }}"; fi
    with_items: "{{ possible_pythons }}"
    register: pystat
    when: pystat is not defined or (pystat.stdout|length) == 0
  - set_fact:
      ansible_python_interpreter: '{{ pystat.results | selectattr("stdout") | map(attribute="stdout") | first }}'
  # NOW we can run the fact gathering step
  - setup:
  - debug:
      msg: and now we are off to the races!

我认为当然也可以使用shell迭代来做到这一点，但是我没有测试它，也没有考虑所有的极端情况：

- raw: |
     PYTHONS='{{ possible_pythons | join(" ") }}'
     for p in $PYTHONS; do
       if [ -x "$p" ]; then echo "$p"; break; fi
     done
  register: the_python

有关ansible_python_interpreter

1 个答案: