node-microservice-boilerplate

A microservice boilerplate for Node.js

A boilerplate to help start Node.js projects quickly and effectively. Packaged with Typescript, Docker, Kubernetes, PM2, Eslint, Prettier, VSCode config, Winston, Typedoc, Nodemon, AVA, PromClient, JWT, Editorconfig, OpenAPI/Swagger, Jaeger/Open Tracing, etc.

Basic Concepts behind the architecture

1. All the code you write resides within src folder and gets built when you start the server in dist folder.
2. All configuration is stored against the respective environment in src/config/env folder and .env file in the root.
3. All your navigation routes are placed in src/app/routes folder. You can place as many js files as you want within.
4. All your logic is written in controllers placed within src/app/controllers folder.
5. If you have utility functions which you will use across controllers, put them in src/app/utils folder.
6. All the server side rendered template files are to be placed in src/app/views folder.
7. All your static files and content is to be placed in the public folder.

Deployment instructions

Normal Installation

1. Install node.js and git
2. Add them to path if not already in path.
3. Install typescript globally with yarn typescript -g
4. Run yarn install to get all the dependencies installed
5. cd to the project directory and run tsc to build using typescript
6. Run yarn start-dev

Cloning & Running with Docker

1. Install Docker and Docker Compose
2. Clone/Download the repository and CD to it
3. Run docker-compose up to get the containers installed and started.
4. Run docker-compose -f docker-compose-prod.yml up to get the containers installed and started in production environment.

Pulling & Running from Docker Hub

1. Pull the container from DockerHub with the command docker pull ajays97/node-microservice-boilerplate (Add version if you want)

2. Then, start the container using the command:
   docker run -d -p 8085:8085 -p 9229:9229 -e NODEJS_PORT=8085 -e NODE_ENV=development -e NODEJS_IP=0.0.0.0 ajays97/node-microservice-boilerplate

Running via PM2

1. Install node.js and pm2
2. CD to the pm2 directory
3. Run pm2 start pm2-dev.json to start the development cluster

Running with Kubernetes

The project comes with some default kubernetes deployment and service yaml configurations. To use it, do kubectl apply -f kubernetes/deployment.yaml and kubectl apply -f kubernetes/service.yaml once Kubernetes and Kubectl are installed onto your system and configured.

Building the code

1. Run yarn build to build the typescript code, copy relevant files

Logging Middleware

Logs can be added by using the log function from error.utils by specifying the log level, payload, SPAN if using Jaeger and a tag object with key-value pairs.

log('info', {
  message: 'Log message here',
  key1: value1,
  key2: value2
});

Use the toggles provided in config to decide where you want to write logs. By default, console logging is enabled and file logging is disabled but this can be changed by using the config.

Static Files

You can place all your static files in the public directory and that will get served by the server directly

Exposing Metrics

The project is bundled with prom-client to enable exporting the metrics to prometheus with ease. You can access the default metrics at “/metrics” endpoint. You can modify what you export in src/app/controllers/metrics.server.controllers.ts

Formatting

The project comes with prettier configs and extensions built in.

You can format the project manually by running the command npm run format and prettier will format the project for you.

You may want to install an extension for your IDE though. More details on the same is available at https://prettier.io

Linting

You can customize rules if needed using the .eslintrc file placed in the root directory. If you are using VSCode, you can have the ESLint extension installed. While linting is run everytime you build/start the server, you can manually run it by npm run lint

Generate API Documentation (OpenAPI/Swagger)

You can generate documentation site by providing details regarding all the endpoints in the apidoc.yaml file in the root and once done, you can start the server and your documentation will get exposed in /api-docs URL. You can use tools like OpenAPI GUI to help generate the YAML file for you.

Generate Code Documentation

You can generate documentation based on your code by running npm run document. Once generated, you can find the documentation in the docs folder. See this to know how to document your code to be rendered by the doc generator.

JWT Authentication

Helpers Utilities for doing JWT authentication is added to src/app/utils/auth.utils.ts and controllers for the same in src/app/controllers/auth.server.controller.ts. The private and public keys are placed in the creds folder. You can generate your own by using tools like this.

No database implementation has been made so far and hence you may need to add your own tables/collections, create users, store secret keys, etc. which is more of an implementation detail.

Editorconfig

A .editorconfig file has been added to the project to enable consistency in development across different IDEs used by different developers. Visit http://editorconfig.org for more information. You might have to install plugins in your editor to get this to work.

Running Tests

You can run your tests by using npm run test command. The project is bundled with AVA as the test framework.

Environmental Variables

You can set up the environmental variables in the .env file and that will get used in files like docker-compose, Dockerfile, etc. In addition to this, you can set up the rest of the environmental configurations in config/env/yourenv.ts where yourenv can be anything and the respective configs will get loaded depending on env you set.

Development vs Production Environment

The development and production environments has some notable differences in their implementation. Everything is handled automatically when run with the right compose file.

1. Use of nodemon in development and directly running with node in production
2. Single Stage Build for Docker image in development and Multi Stage Build for Docker image in production
3. src mounted as volume in development and no volume mount in production
4. All files copied to container in development and only necessary files copied to container in production
5. devDependencies installed in development but devDependencies ignored in production

Compatibility

Since this project uses all the latest features of the node ecosystem, it requires Node >= v10.0.0

Contributors

1. Ajay Srinivas