Contributing to GradeJS Backend

Before doing the further actions, install Node.js and Yarn, clone this repository, open a terminal in the root of the repository and run:

yarn install

The required Node.js version is specified in the package.json file. Consider using NVM utility to switch to the correct version.

Running locally

For some unit tests, such as API route testing, you will need to use a local PostgreSQL instance. A local database is also needed for running the application locally.

Create a docker container with a local PostgreSQL instance by running this commands:

docker-compose up -d db

# Create all required tables.
yarn --cwd packages/shared migration:run

# Run later to stop the database service:
docker-compose down

Run the server API:

yarn --cwd packages/public-api build
yarn --cwd packages/public-api start

Run the web application:

yarn --cwd packages/web start

In order to view database logs, consider enabling TypeORM logging in the connection.ts file at the @gradejs/public-api package or using the following command:

docker logs -f db

Server Deployment

Initial setup

Sign in to AWS.
Consider creating a PostgreSQL table before the first deployment:

Open the AWS RDS page.
In the database list click on the Create Database button.
Select the following parameters: Standart create, Amazon Aurora PostgreSQL-Compatible Edition version 13, Production template.
Enter the database name, for example gradejs-public.
Select a burstable DB instance class: db.t3.medium
Select Don't create an Aurora Replica option.
Consider disabling public access for the production environment.
In the Additional configuration section set the Initial database name to gradejs.
Click in the create button. The initial migration will be executed by the application during the first deployment.

Create an AWS Elastic Beanstalk application:

Open the AWS Elastic Beanstalk page, select the Applications section and click on the Create a new application button.
Enter or select the application name, for example gradejs.

Create Amazon SQS worker queues:

Open the page and click on the Create queue button.
Select the standart queue option and specify the name gradejs-public.
Set the visibility timeout to 2 minutes.
Click on the Create queue button.

Create an AWS Elastic Beanstalk API environment:

Open the AWS Elastic Beanstalk page and click on the Create a new environment button.
Select Web server environment option.
Type the newly created application name (gradejs).
Specify an environment name, for example gradejs-public-api. Leave the domain and description name fileds blank.
Select the latest version of Node.js 16 managed platform on Amazon Linux 2. The version minimal required version is 5.5.1.
Use the sample application code and click on the Configure more options button.
Select the Custom configuration configuration preset.
In the Software section click on the Edit button.
Enable the CloudWatch log streaming with a 5 day retention.
Specify the AWS_REGION environment variable. It should be the same as your current AWS environemnt. For example, us-east-2.
Specify the DB_URL environment variable. It should point to the newly created RDS database. For example: postgres://postgres:<secret>@<hostname>/gradejs.
Specify the INTERNAL_API_ORIGIN environment variable. Consider using our staging internal API origin: http://fpjs-dev-gradejs-internal-api.eba-fybi4md5.us-east-1.elasticbeanstalk.com
Specify the SQS_WORKER_QUEUE_URL queue. It should be a newly created gradejs-backend queue. For example https://sqs.us-east-2.amazonaws.com/<account_id>/gradejs-backend.
Specify the AWS_REGION environment variable. It should be the same as your current AWS environemnt. For example, us-east-2.
Specify the EB_START environment variable: api.
Click on the Save button and then click on the Create environment button.

Create an AWS Elastic Beanstalk worker environment:

Select Worker environment option.
Type the newly created application name (gradejs-public).
Specify an environment name: gradejs-public-worker. Leave the domain and description name fileds blank.
Select the latest version of Node.js 16 managed platform on Amazon Linux 2. The version minimal required version is 5.5.1.
Use the sample application code and click on the Configure more options button.
Select the Custom configuration configuration preset.
In the Software section click on the Edit button.
Enable the CloudWatch log streaming with a 5 day retention.
Set the same AWS_REGION, DB_URL, SQS_WORKER_QUEUE_URL environment variables.
Set the EB_START environment variable to worker.
Click on the Save button and then click on the Create environment button.

Setting up a Continuous Deploy pipeline:

Go to the AWS CodePipeline page and click on the Create pipeline page.
Type a pipeline name gradejs-public@production-deploy and click on the Next button.
Select the Github 2 source provider and select the gradejs repository. For the production environment a branch name should be master and for staging -- develop.
Click on the Next button.
Select the AWS CodeBuild build provider.
Create a AWS CodeBuild project. The name of a project should be gradejs-public. Select the Ubuntu operating system with the latest version of the aws/codebuild/standart:5.0 image. Click on the privileged flag. Leave the rest as defaults and click on the Continue to CodePipeline button.
Click on the Next button on the build stage page.
Select the AWS Elastic Beanstalk deploy provider.
Select the gradejs-public-api environment name.
Click on Next button and then on the Create pipeline button.
Click on the Edit button in the newly created pipeline. Edit the Deploy stage and click on the Add action group button
Type the deploy-public-worker action name and select the AWS CodeBuild build provider.
Select the BuildArtifact as input artifact.
Select the gradejs-public-worker environment name.
Click on Done button and then on Save button.