Skip to content

Latest commit

 

History

History
123 lines (73 loc) · 3.45 KB

README.md

File metadata and controls

123 lines (73 loc) · 3.45 KB
Raw

Fullstack docker compose

This compose example will provide:

  • front-end: nginx server to
    • serve frontend assets
    • forward requests to the backend
  • back-end: Postgrest backend
  • database: PostgreSQL server

Architecture schema

Structure

├── db
│   ├── data             # Database data (need root rights to see files)
│   ├── init             # Custom scripts to be run on the 1st DB startup
│   └── runtime          # Custom migrations
├── docker-compose.yml   # Structure the whole application
├── nginx
│   ├── certs            # SSL certificates for localhost
│   └── nginx.conf
├── web                  # Web project root path
│   └── dist             # Build path used by nginx to serve web assets
└── README.md

Exposes

  • http://localhost redirecting automatically to 443 port
  • https://localhost with following routes:
    • /api: postgrest API on /api path
    • / all assets available in web/dist directory

Configuration

Create and install certificate

cd nginx/cert
mkcert localhost
mkcert install

Start the stack

POSTGRES_PASSWORD=AZERTY docker-compose up

How to?

List APIs

Postgrest API are here: https://localhost/api

Add a web project

You can use the web framework of your choice.

The only thing you need is to put your built web assets into the web/dist directory.

Example with vite and react

Here we show how to integrate a reactJS project using viteJs:

 # Create the project
 yarn create vite web --template react
 
 # Build it
 cd web
 
 # Download dependencies
 yarn

You can use the dev mode and have a server outside of the docker fullstack with:

yarn dev # exposes http://localhost:3000

Or you can build the web application to let nginx serve the assets:

yarn build # access your application through https://localhost

Don't forget to rerun yarn build every time you want to update the assets for nginx.

Connect to database

You can connect a psql client to your db:

docker exec -it docker-fullstack_db_1 psql -h localhost postgres postgres

Add a SQL script to be run on the first database start

Put your sql script in db/init directory

Run a SQL script from db/shared

Place your myscript.sql file in db/runtime and run:

docker exec -it docker-fullstack_db_1 psql -h localhost postgres postgres -f /runtime-scripts/myscript.sql

Reset totally the database

sudo rm -rf db/data/pgdata

Step by step example

Create a migration file db/runtime/books.sql with:

CREATE TABLE IF NOT EXISTS books (title text);

Run this migration file against the database:

docker exec -it docker-fullstack_db_1 psql -h localhost postgres postgres -f /runtime-scripts/books.sql

Postgrest API is now up to date, and contains the new books entity: https://localhost/api

But for now, there are no books: https://localhost/api/books

You can create a book with:

curl -X POST https://localhost/api/books -H "Content-Type: application/json" -d '{"title": "My title"}'

You can list the new book with: https://localhost/api/books

Documentation