有没有办法在环绕Ansible playbook中的角色时设置环境变量？

Question

I need loop over an Ansible role and save the index of the iteration.

My goal is to use the number(which is INDEX in this case) in my-role for each iteration. My-role executes other playbooks and I need the value of the INDEX for each iteration. I want to use that index in another ansible-playbook. So my plan is to set an environment variable to read from the other plaaybook.

This is my code:

- name: my-role
  with_sequence: 'start=0 end={{ number_of_users|int }}'
  include_role:
    name: my-role
  vars:
    user_name: '{{ user_temp[item|int] }}'
    queue_name: '{{ queue_temp[item|int] }}'
  lineinfile:
    dest: "/etc/environment"
    state: present
    line: 'export INDEX=[item|int]'

Apparently I cannot do two statements at the same time.

The error message is:

ERROR! conflicting action statements: include_role, lineinfile

Is there a way to set an environment variable while looping over a role?

Answer 1

Is there a way to set an environment variable while looping over a role?

Yes, using the apply: option of include_role:

- with_sequence: start=0 end=3
  include_role:
    name: my-role
    apply:
      environment:
        INDEX: '[{{item}}]'

BTW, even if ansible had let you run lineinfile alongside that role, just putting an entry in /etc/environment is highly unlikely to automatically expose that environment variable to the role. It would require that every ssh connection made for every task actually source /etc/environment , which might happen but it's unwise to count on it when ansible offers you an environment: directive that is explicitly designed to do that.