johnzech/self-destruct

testing testing

bento-box

A python web developer's box of goodies.

Table of Contents

• Setup
  • Set up the Vagrant Box
  • SSH to the machine
  • Putty setup (optional)
  • Install on AWS, Google, Digital Ocean, etc
• Get Started
  • Vagrant up gets you...
  • Creating a new module
  • Creating a new db schema
  • Creating a new cron job
• Architecture
• Note on db versioning/migrations
• License

Setup

Set Up the Vagrant box

1. get the deps:
   • Vagrant
   • Virtualbox
   • optional: winscp, putty, puttygen
2. open terminal/cmd in this dir
3. vagrant up

SSH to the machine

2 options:

• vagrant ssh
• your ssh client of choice

Putty setup (optional)

TO USE PUTTY, convert the autogenerated private key to ppk

• autogenerated: ./.vagrant/machines/default/virtualbox/private_key (pem)
• putty: ./.vagrant/machines/default/virtualbox/private.ppk (putty) To Convert:

1. run generate_putty_key.bat after 'vagrant up' completes
   • NOTE: you'll need winscp in your path for this step to work
   • ALTERNATIVELY: convert the key manually using puttygen
2. configure putty session:
   • user: vagrant
   • host: localhost
   • port: 2222
   • private key: ./.vagrant/machines/default/virtualbox/private.ppk

Install on AWS, Google, Digital Ocean, etc

1. make sure your instance is running almalinux 8.x
2. clone repo in /var/www/modules
3. run install.sh

Get Started

What you get

vagrant up

Gets you...

• almalinux
• all rpms updated
• python38 installed
• fastapi installed
• hypercorn installed
• postgres 13 and postgis 3 installed
  • also a starter database called 'bento' and a schema called test
• migra installed
• empty 'root_app' module listening on the '/' path
• hello world fastapi app installed and configured in nginx
• example api module installed (with an actual database schema behind it)

Creating a new module

**for the sake of example, we'll act like we're adding a module named 'carbs'

1. create a new dir in the root of this repo (/carbs/)
2. make sure it has the following (probably best to copy one of the examples)
   • __init__.py (empty - if you're reading this in an editor... otherwise known as init.py)
   • carbs.py (main fastapi app here)
   • carbs.toml ( hypercorn config items)
   • carbs.service (systemd service file)
   • carbs_conf_d.conf (nginx config file (conf.d dir) - loaded outside of the server directive)
   • carbs.conf (nginx config file (default.d dir) - loaded inside of the server directive)
   • ./templates/ dir (optional, if your module has a web ui) **If copy/pasting, make sure to review each of the above files
3. add a line to /provision/modules.sh for your new carbs module
   • will run next time your vagrant box provisions
   • to run the provision step manually (just for your module), comment out accordingly and run modules.sh
   • this step:
     • installs your module as a uwsgi proc running as a service
     • adds the nginx config to the appropriate dir
   • systemctl command will match the filename of your .service file (systemctl start carbs)
4. restart nginx
   • 'systemctl restart nginx

Creating a new db schema

**for the sake of example, we'll act like we're adding a schema named 'carbs'

1. add a schema create file in the db dir (/db/carbs_schema.sql)
   • this file should start with 'CREATE SCHEMA carbs;'
2. add a line to /provision/schemas.sh
   • runuser -l postgres -c "psql -U bento -f /var/www/modules/db/carbs_schema.sql"
   • note: this will install the carbs schema to the bento database - if you've created your own database, use that

Creating a new cron job

**for the sake of example, we'll act like we're adding a cron named 'carbs'

1. create the new dir in /crons/carbs
2. make sure it has the following (probably best to copy the example_cron)
   • carbs.logrotate
   • carbs.py (main python cron job)
3. add a line to /crons/crontab (make sure that there is a newline at the end of the file)
   • in our case */1 * * * * cd /var/www/modules/crons/carbs && /bin/python3 /var/www/modules/crons/carbs/carbs.py >> /var/log/carbs/carbs.log
4. add a line to /provision/crons.sh
   • crons+=(carbs)

Architecture

The following bash output is the directory structure and organization of bento-box:

tree

.
├── config.py
├── crons
│   ├── crontab
│   └── example_cron
│       ├── example_cron.logrotate
│       ├── example_cron.py
│       └── say_hello.py
├── db
│   ├── create_database.sql
│   └── test_schema.sql
├── example_api
│   ├── example_api.conf
│   ├── example_api_conf_d.conf
│   ├── example_api.py
│   ├── example_api.service
│   ├── example_api.toml
│   └── __init__.py
├── generate_putty_key.bat
├── hello_world
│   ├── hello_world.conf
│   ├── hello_world_conf_d.conf
│   ├── hello_world.py
│   ├── hello_world.service
│   ├── hello_world.toml
│   ├── __init__.py
│   └── templates
│       └── hello_world.html
├── __init__.py
├── install.sh
├── LICENSE
├── provision
│   ├── crons.sh
│   ├── dev_env.sh
│   ├── enable_ssl.sh
│   ├── modules.sh
│   ├── nginx_conf
│   ├── nginx.sh
│   ├── packages.sh
│   ├── postgres.sh
│   ├── schemas.sh
│   └── static.conf
├── README.md
├── root_app
│   ├── __init__.py
│   ├── root_app.conf
│   ├── root_app_conf_d.conf
│   ├── root_app.py
│   ├── root_app.service
│   ├── root_app.toml
│   ├── templates
│   │   ├── base_template.html
│   │   └── index.html
│   └── wsgi.py
├── static
│   ├── css
│   │   └── root.css
│   ├── favicon.ico
│   └── image
│       ├── Bento.png
│       └── title_logo.png
├── utils
│   ├── db.py
│   ├── __init__.py
│   └── __pycache__
│       ├── db.cpython-36.pyc
│       ├── db.cpython-38.pyc
│       ├── __init__.cpython-36.pyc
│       └── __init__.cpython-38.pyc
└── Vagrantfile
14 directories, 55 files

Note on db versioning/migrations

bento-box does not solve for db versioning at the moment. But it does include migra, a db diff tool. That's a good start toward a sane db "non-versioning" scheme. Read up on it here: https://djrobstep.com/docs/migra/quickstart

License

MIT © John Zechlin