node-graphql-boilerplate

This repository contains my best practices, libraries setups, boilerplate examples and basic configurations for nodejs backend based on these technologies:

• Code
  • Typescript
  • GraphQL
  • custom Graphql Typescript framework
  • Type-orm
  • environment variable parsing
  • sending emails via ethereal or AWS SES
• Security
  • oAuth2 (google authorization)
  • local bearer authorization
• Database
  • Postgres
• devops
  • Docker
  • AWS serverless monorepo deployment
  • add multiple deployment environments (test|prod|test|etc...)
  • terraformed infrastructure
  • extract TF backend state into s3 + add dynamo locking

TODO:

• Create custom GQL ID type with prefix:UUID
• AWS
  • test AWS SES
  • add xray AWS lambda tracing
• apollo tracking does not work after webpack build
• add FB oAuth2
• tests
• sql migrations
• check the keycloak support
• add pgadmin into docker-compose?

Requirements

• nodejs14+
• docker
• docker-compose

Configuration

Setup RDS (or whatever you want) by hand (RDS setup is not included in the .tf files)

Copy .env.example file into .env and setup your local variables.

Environment details are described in the /src/appConfig.ts.

Database docker container environments are setted in the /docker-compose.yml.

Installation

All npm scripts are available also with the docker: prefix to call them directly in the running container

with docker

1. run npm install
2. run npm run docker:db:hard-init - to setup database into init state
3. run npm run docker:dev - to start app and database

without docker

1. run npm install
2. run npm run db:hard-init - to setup database into init state
3. run npm run dev - to start app and database

Helper script

kill $(lsof -ti:2020)

steps to create this boilerplate into new project

• Create SQL database and get Credentials
• Create dynamo table
• Setup .env(stage-1|production) file
• change AWS S3 Terraform backend
• Setup tfvars.(stage-1|production) files
• deploy infrastructure

How to run & reconfigure this fullstack repository

Run local development

git clone .....

setup local env variables

# setup envs
cp .example.env .env
cp .example.env .env
cp .example.env .env

# install dependencies
npm install

change database name + change db_name in docker-compose.yml

# run database:
npm run docker:db:up

# add init data
npm run docker:db:hard-init

deployment

Postgre database

In AWS console create new RDS and set it public with proper settings of security groups.

Then connect into this created instance and add a new user. you can use scripts like this:

CREATE DATABASE my_app_production

CREATE USER my_app_production_user WITH PASSWORD 'my-secret-password';

GRANT ALL PRIVILEGES ON DATABASE "my_app_production" to my_app_production_user;

-- in scope of newly created db create uuid extension
CREATE EXTENSION IF NOT EXISTS "uuid-ossp";

add pg user tutorial

DynamoDB database

Go to AWS console and create new dynamo table with PK named: PK and SK named: SK

setup .env.(production|stage-1).env files which will be deployed with your code into the AWS lambda

cp .example.env .env
cp .example.env .env.stage-1.env
cp .example.env .env.production

now you should setup

• POSTGRES_USER
• POSTGRES_PASSWORD
• POSTGRES_DB_NAME
• POSTGRES_HOST
• AWS_ACCESS_KEY_ID -> this should be removed in the near future
• AWS_SECRET_ACCESS_KEY -> this should be removed in the near future
• AWS_DYNAMO_LOG_TABLE_NAME=node-boilerplate-stage-1
• ...

terraform state

go to init-terraform-backend/setup-backend.tf and change name of your dynamodb table + s3 bucket

cd init-terraform-backend

terraform init
terraform apply

then set your newly created resources into /terraform/main > terraform > backend "s3" > bucket + dynamodb_table

Setup Terraform variables

Don't forget you have to be logged into AWS CLI from PC which deploy your terraform infrastructure

go to /terraform folder and change globalLocals.tf variables

Now you should deploy both your environments with just simple scripts:

terraform init

npm run deploy:infrastructure # or d:i
npm run deploy:code:production # or d:c:p
npm run deploy:code:stage-1 # or d:c:s

rename deploy scripts

by name of your lambdas change .sh scripts in the deploy-scripts folder