deploymentutils 0.3.0

Small python package to facilitate deployment of some personal projects.

deploymentutils

This repo contains a small python package to facilitate deployment of some personal projects.

Overview

This package provides a thin layer on top of fabric to execute commands with a state like

• current working directory
• activated virtual environment (not yet implemented)

It also tries to simplify to deploy/maintain multiple instances of the same software but with varying fixtures, including one or more local instances for testing.

Motivation

The package is mainly intended to facilitate deployment tasks (e.g. for django apps) by running a simple python script. Compared to configuration management tools like Ansible this approach is far less powerful and scalable. However, it might be easier to understand for developers and thus lowering the hurdle to deploy applications by them selves.

Assumed Directory Layout

The scripts marked with [*] make use of this module.

<general project dir>
│
│
├── project-src-repo/                       ← official project repo
│   ├── .git/
│   ├── deployment/
│   │   ├── some_files/
│   │   ├── deploy.py      [*]              ← original deployment script  (delivered by the project)
│   │   └── ...
│   ├── django_project/
│   │   ├── settings.py
│   │   └── ...
│   ├── django_app/
│   │   ├── views.py
│   │   └── ...
│   ├── manage.py
│   └── ...
│
├── instance-specific/
│   ├── README.md
│   ├── demo
│   │   ├── .git/
│   │   ├── deploy.py
│   │   ├── project_data
│   │   └── ...
│   ├── production
│   │   ├── .git/
│   │   ├── deploy.py      [*]
│   │   ├── project_data
│   │   └── ...
│   ├── testing
│   │   ├── .git/
│   │   ├── deploy.py      [*]
│   │   ├── project_data
│   │   └── ...
│   └── ...
│
├── local_testing/               ← contents of this directory are autogenerated
│   ├── deploy.py          [*]
│   ├── <appname>_deployment
│   └── ...
.

Status

Still under development and not comprehensively tested.

Known Issues

• If a command started by c.run("some_command") is reading input, then the calling python process waits 'forever', i.e. until interrupted manually.
  • possible solution fragment: https://stackoverflow.com/questions/35751295/python-subprocess-check-to-see-if-the-executed-script-is-asking-for-user-input