Skip to content

Latest commit

 

History

History
86 lines (73 loc) · 2.81 KB

provisioning-servers-for-hosting.md

File metadata and controls

86 lines (73 loc) · 2.81 KB
Raw

Provisioning webserver

In this guide we'll explain how to use the included provisioning script to install a Pipit generated application on a server.

Requirements (webserver, aka ansible host)

The server should have these applications/packages installed:

  • Linux (Ubuntu 20.04+ is preffered)
  • Nginx
  • uWSGI
  • Python 3.8+
  • PostgreSQL 12+
  • PostGIS for PostgreSQL
  • GDAL (required for PostGIS)
  • Node 12+
  • PM2
  • psycopg2-binary (this is required for the provision script that will create db and users)

Linux should have these users (with passwordless login using RSA keys):

  • "root" - Used when provisioning web server
  • "deploy" - Used for deployment

Configuration:

  • Systemd jobs for Nginx, uWSGI and pm2
  • Nginx configuration are stored at /mnt/persist/nginx/conf.d/*
  • Node configuration are stored at /mnt/persist/nodejs/*
  • Your web applications are stored at /mnt/persist/www/*
  • The root user can access the psql shell without password

Requirements (your computer, aka control node)

  • A fully generated Pipit project
  • Rsync installed
  • A MacOS or Linux computer (Ansible does not support Windows)
  • Access to the webserver over ssh with both the "root" and "deploy" user

Guide

  • Begin by going to the deploy dir in your project and install Ansible

    >>> cd deploy
>>> python3 -m venv venv
>>> . venv/bin/activate
>>> pip install -r requirements.txt

  • Make sure you can connect to the server by pinging it

    >>> ansible -i stages/stage.yml webservers -m ping
stage1 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python3"
    },
    "changed": false,
    "ping": "pong"
}

  • After this, we install Ansistrano:

    >>> ansible-galaxy install -r requirements.yml

  • Now that we have everything installed, lets run the provisioning:

  • For stage

    >>> ansible-playbook provision.yml -i stages/stage.yml

  • For prod

    >>> ansible-playbook provision.yml -i stages/prod.yml

  • This will script will run the necessary steps to make sure your application is ready to be deployed

  • The next step is to run a deploy to sent our application to the server and start it

  • For stage

    >>> ansible-playbook deploy.yml -i stages/stage.yml

  • For prod

    >>> ansible-playbook deploy.yml -i stages/prod.yml

  • Done!

Note

  • This script does not perform server provisioning, only application provisioning. But there are plenty of guides on how to do this.