ArianBahrampour / expressjs-production

Express js API

This is a simple express js api with typescript and mongodb.

Up and Running

To start project on production environment using docker use below steps.

• Run cp .env.example .env to copy .env.example contents into .env file.
• Change environment variables to appropriate values :
  • PORT
  • DBURL
  • MONGO_DB_ROOT_USERNAME, MONGO_DB_ROOT_PASSWORD, MONGO_DB_NAME
  • Be sure to fill mongodb credentials in DBURL variable.
  • mongodb url format is mongodb://<username>:<password>@xcoin-mongo:<port>/<database>
• Run docker-compose up --build - d and wait until docker download images (for the first time)
• Open http://127.0.0.1:{PORT} in your browser to check service
• Run docker-compose down to stop containers

Database migrations

You should enter exposed db url in order to run seeds or migrations. Another way is to use docker-compose exec api bash to enter container and run migrations.

Then start the migrations using following:

Seeds

npm install -g ts-node

ts-node src/scripts/seed.ts

My steps for code review

Docker

First of all I started writing docker file and docker compose for the project to run it in a containerized environment. I used the official node image and added the necessary dependencies. I also added the mongo image to the docker compose file. Mongo db data is stored in a volume so no data will be lost.

• There is an nginx.conf file in the project which can used for proxing the api.

Environment variables

After that I added the environment variables to the docker compose file and also added the .env.example file to the project. I also added the dotenv package to the project to read the environment variables from the .env file.

Logs And Error Handling

I added the winston package to the project for logging. I created a logger.ts file in the project and added the necessary configurations. I also added the express-async-errors package to the project for error handling. I created an error.ts file in the project and added the necessary error classes. Each request is logged with the winston package. I also added the morgan package to the project for logging the requests.

API Validation and Testing

APIs needed validation for body with propper response. Also added some test for the project. I used the jest package for testing. I created a test folder in the project and added the necessary tests. I also added the supertest package to the project for testing the APIs.