nicovogel / node-ms

Compare different communication factors in a nodejs microservice environment

Microservice Communication evalutaion

This project aim to compare multiple aspects which are related to nodejs microservice communication. The main project can be found inside the lan folder, while the prototypes are located inside prototype.

LAN Party Application

Prerequisit

Before starting, you need to eigther create a .env file within the lan folder or change the following environment variable inside the docker-compose.yml file.

Set the domain name of the server as value for SERVER_NAME inside .env or docker-compose.yml. If you do not have a domain, use localhost.

Startup

To start the application navigate into the lan folder and use docker-compose up to startup the project.

URL's

Everythin is accassible through nginx. Thus the following urls are available.

domain equals the provided SERVER_NAME value

url	service
domain	angular website
domain/api/account	account service rest endpoint
domain/api/event	event service rest endpoint
domain/api/billing	billing service rest endpoint
mgnt.domain/metric	grafana ui
mgnt.domain/mongo:8080	mongo instance (used to connect with compass)
mgnt.domain/rabbitmq	rabbitmq management ui

User and password overview for the management services (can be configered via docker-compose.yml)

service	user	password
grafana	admin	admin
rabbitmq	guest	guest
mongodb for account	account	account
mongodb for event	event	event
mongodb for billing	billing	billing

Debug options

Each service and the angular app feature a live reload support. This can be enabled via docker-compose.yml.

For the services (e.g. account, event, billing) uncomment the environment variable NODE_ENV and the volume mount.

Example:

event:
    build: event
    environment: 
      NODE_ENV: development
    volumes:
      - ./event/src:/app/src

For the angular app uncomment the environment variable DEPLOY_ENV for nginx.

Example:

nginx:
    build: nginx
    ports: 
      - 80:80
      - 8080:8080
    depends_on:
      - grafana
    environment: 
      PORT: 80
      MONGO_PORT: 8080
      DEPLOY_ENV: development
      SERVER_NAME: ${SERVER_NAME}

And also navigat into the angular code (lan/nginx/app) and execute:

npm run install
npm run debug

Mongo Databases

There is only one mongodb instance, but it contains several databases, in fact one for each service. The database users are created within the lan/mongo-init.js which is mounted into mongo.

Database names for each service: |db|service| |node-ms-account| account| |node-ms-event| event| |node-ms-billing| billing|

Mongodb is also accessible via mgnt.domain/mongo. It is archived by using nginx upstream module. Thus, it uses the port 8080.

Nginx

Normaly nginx does not support environment variables in its configuration. Therfore, a entrypoint script overrides placeholders within the nginx.config.

The scripts are located at lan/nginx/scripts.

Endpoint definitions

A gerneral overview is available within the lan/shared folder. The file messages.ts contains all rabbitmq events and there data structure and rest.ts shows the endpoints and there required data to use them.

Blank Service

There is a blank micro service located at lan/_blankService. It contains all neccessary components and is ready to use.

Dev Container

To abstract the development environment and ensure the same developer experience on each machine Dev Container is used. The extension is called Remote - Containers. There is a good explanation for the extension in the vs code docs: Developing inside a Container

Reqiements to startup the project:

• Docker running
• VS Code running
• Remote - Containers extension

Steps:

1. Open the cloned repo in VS Code
2. Select Remote-Container: Reopen in Container in the VS Code action bar

This will build the Dockerfile and show all files in your VS Code as soon as the build is finished. Now the dev environment is setup.