uttamsutariya/multi-containerized-backend

Complete docker multi container setup for node.js (express) server application with docker-compose

Multi container backend with node/express, redis, mongodb & docker

Application architecture

Server folder architecture

|- app
    |- index.js
    |- controllers/
    |- middlewares/
    |- routes/
    |- models/
    |- config/
    |- nginx/
    |- package.json
    |- package-lock.json
    |- Dockerfile
    |- docker-compose.yml
    |- docker-compose.prod.yml
    |- docker-compose.dev.yml
    |- docker-compose.backup.yml
    |- .gitignore
    |- .dockerignore
|- docker-compose-down.sh
|- docker-compose-up.sh

To start a application

./docker-compose-up.sh

To stop application

./docker-compose-down.sh

Content of docker-compose-up.sh

docker compose --env-file .env -f docker-compose.yml -f docker-compose.prod.yml up -d --build

Content of docker-compose-down.sh

docker compose -f docker-compose.yml -f docker-compose.prod.yml down

To re-build a specific image on production (eg. mongodb & nginx images may not be needed to re-build every-time application-logic changes...)

docker compose --env-file .env -f docker-compose.yml -f docker-compose.prod.yml up -d --build backend

here image that must be re-build is backend

we can even pass --no-deps flag to not re-build the image on that the our changing image depends

Current production workflow

Developer build image on dev server it-self & push it to dockerhub, then production server pull that image from dockerhub & run docker compose up and starts all containers. By building image on dev-server, it reduce compute power of production server which is good for application performance

Previous production workflow

Developer push changes to github, then production server takes git pull & re-build the image which takes more compute power of production server which is not good for application. After re-building it run docker compose up and starts all containers.

To install docker on ubuntu server

go to https://get.docker.com/