LSO: an API that allows for remotely executing Ansible playbooks.

Code documentation

Code documentation can be found at https://workfloworchestrator.org/lso

Quick start

This is a quick setup guide for running on your local machine.

As a Docker container

To run LSO as a Docker container, build an image using the Dockerfile.example as an example. Be sure to update requirements.txt and ansible-galaxy-requirements.yaml accordingly, depending on your specific Ansible collection and -role needs.

Use the Docker image to then spin up an environment. An example Docker compose file is presented below:

services:
  lso:
    image: my-lso:latest
    env_file: 
      .env  # Load default environment variables from the .env file
    volumes:
      - "/home/user/ansible_inventory:/opt/ansible_inventory:ro"
      - "~/.ssh/id_ed25519.pub:/root/.ssh/id_ed25519.pub:ro"
      - "~/.ssh/id_ed25519:/root/.ssh/id_ed25519:ro"

This will expose the API on port 8000. The container requires some more files to be mounted:

• An .env file: Sets default environment variables, like ANSIBLE_PLAYBOOKS_ROOT_DIR for the location of Ansible playbooks inside the container.
• Environment variables: Specific configurations, such as ANSIBLE_ROLES_PATH, can be directly set in the environment section. This is ideal for values you may want to override without modifying the .env file.
• An Ansible inventory for all host and group variables that are used in the playbooks
• A public/private key pair for SSH authentication on external machines that are targeted by Ansible playbooks.
• Any Ansible-specific configuration (such as collections_path, roles_path, etc.) should be set using environment variables. ANSIBLE_ROLES_PATH is given as an example in the Docker compose snippet above.

Install the module

As an alternative, below are a set of instructions for installing and running LSO directly on a machine.

One of these should be what you're looking for:

• Install the latest release

  python3 -m venv my-venv-directory
  . my-venv-directory/bin/activate

  pip install orchestrator-lso

• Install the source code

  git clone https://github.com/workfloworchestrator/lso.git && cd lso
  python3 -m venv my-venv-directory
  . my-venv-directory/bin/activate
  
  pip install flit
  flit install --deps production
  
  # Or, for the full development environment
  flit install --deps develop

Running the app

• Set required environment variables; see env.example for reference.
• If necessary, set the environment variable ANSIBLE_HOME to a custom path.
• Run the app like this (app.py starts the server on port 44444):

  source .env && python -m lso.app

Task Execution Options

1. Celery (Distributed Execution)

• For distributed task execution, set EXECUTOR=celery.
• Add Celery config in your environment variables:

CELERY_BROKER_URL=redis://localhost:6379/0
CELERY_RESULT_BACKEND=redis://localhost:6379/0
WORKER_QUEUE_NAME=lso-worker-queue # default value is None so you don't need this by default.

• Start a Celery worker:

celery -A lso.worker worker --loglevel=info -Q lso-worker-queue

2. ThreadPoolExecutor (Local Execution)

For local concurrent tasks, set EXECUTOR=threadpool and configure MAX_THREAD_POOL_WORKERS.