如何将AWX项目克隆路径传入Ansible Playbook以执行Docker操作？

Great question! When working with Ansible AWX (now part of Ansible Automation Platform Controller), there's a clean, maintainable way to grab your project's cloned path without hardcoding it—perfect for your Docker build step.

The Solution: Use Ansible's Built-in {{ playbook_dir }} Variable #

AWX clones your Git SCM project into /var/lib/awx/project by default, and Ansible has a magic built-in variable called playbook_dir that points directly to the directory where your running Playbook is stored. That's exactly the path you need!

You can use this variable in two ways for your Docker build: either specify it as the build context directly, or set it as the working directory for the command. Also, a quick heads-up: your original docker run uses -it, which will fail in AWX's non-interactive environment—swap that for -d to run the container in the background instead.

Here's your updated Playbook:

---
- hosts: all
  sudo: yes
  remote_user: ubuntu
  gather_facts: no
  tasks:
    - name : Build test-api Docker image
      become: yes
      become_user: root
      # Option 1: Explicitly pass the project path as build context
      command : docker build -t "test-api" "{{ playbook_dir }}"

      # Option 2: Set the working directory, so "." refers to the project path
      # command : docker build -t "test-api" .
      # args:
      #   chdir: "{{ playbook_dir }}"

    - name: Run test-api container in background
      become: yes
      become_user: root
      command : docker run -d -p 80:9001 --name api test-api

Extra Tips for Reliability #

• Avoid hardcoding paths! Hardcoding /var/lib/awx/project works now, but if your AWX setup ever changes (like a custom project storage path), your Playbook breaks. Using playbook_dir makes it environment-agnostic.
• Ditch -it for non-interactive environments AWX runs Playbooks without an interactive terminal, so -it will cause your container to crash immediately. -d runs it detached, which is the right fit for automation.
• Consider using dedicated Docker modules (optional) If you install the community.docker collection, using modules like docker_image and docker_container is more robust than raw command calls—they handle state checks, error handling, and idempotency better. Here's how that would look:

  - name: Build Docker image with dedicated module
     community.docker.docker_image:
       name: test-api
       build:
         path: "{{ playbook_dir }}"
       state: present
  
   - name: Run Docker container with dedicated module
     community.docker.docker_container:
       name: api
       image: test-api
       ports:
         - "80:9001"
       state: started
       detach: yes
  

内容的提问来源于stack exchange，提问作者Frodo