backend

Peerprep Backend

Development Setup (uncontainerized)

1. Ensure that you have PostgreSQL, MongoDB, Redis and node v18 installed.
2. Ensure that you have yarn v1.22.19 installed (e.g. npm i -g yarn).
3. Start PostgreSQL, MongoDB and Redis (e.g. brew services start redis if installed via Homebrew on MacOS)
4. cd to the backend directory.
5. Run yarn install.
6. Copy the .env.example file as .env (e.g. cp .env.example .env) and update the variables where necessary (you may find the Environment Variables setup section below useful).
7. Setup and populate the databases via yarn migrate-and-seed:all.
8. Open multiple terminals to the backend directory, and run each microservice in 'watch' mode (e.g. yarn start:dev http-gateway, yarn start:dev ws-gateway, yarn start:dev question).

   NOTE: The complete list of microservices are the directory names under the backend/apps directory. All microservices have to be started for the backend to run correctly.

Building locally with Docker

The teaching team can choose to setup the dockerized backend locally as follows:

1. Install Docker Desktop if you have not already done so.
2. Copy the .env.docker.example file as .env.docker (e.g. cp .env.docker.example .env.docker) and update the remaining empty secrets (you may find the Environment Variables setup section below useful).
3. In the backend directory, run yarn start:docker to build and run all Dockerized microservices in detached mode locally.
4. Run yarn stop:docker to stop the microservice containers.

Testing (local)

1. Install Docker Desktop if you have not already done so.
2. In the backend directory, run yarn test to run the E2E tests

Database migrations

1. Create a knex (SQL) database migration file via yarn migrate:sql:make {microservice_name} {migration_file_name}.
2. To rollback a knex migration, run yarn knex {microservice_name} migrate:rollback.

Making Commits

1. Run yarn lint and yarn format in the backend directory to lint and format the backend codebase.

Upgrading of user to maintainer role

1. To test out CRUD operations with questions, update the role_id column of the you want entry in the user_profiles table in the peer-prep-user-service database. (e.g. UPDATE user_profiles SET role_id=1 WHERE name='Your Name')

Environment Variables setup

PostgreSQL Variables

• For local PostgreSQL, {MICROSERVICE}_SERVICE_SQL_DATABASE_HOST can be left blank
• Update the USER and PASSWORD related variables

Chatbot API

• Populate CHATBOT_SERVICE_OPENAI_API_KEY={Your own API key from open AI}

JWT Tokens

• Set ACCESS_TOKEN_SECRET and REFRESH_TOKEN_SECRET to some random secret string
• Set the corresponding expiries to something desired (e.g. 1h or 30s or 7d)

Google OAuth

• From Google Console, generate a set of OAuth credentials and put it under OAUTH_GOOGLE_ID and OAUTH_GOOGLE_SECRET
• For local development:
  • Under Authorized JavaScript origins, add http://localhost:4000
  • Under Authorized Redirect URIs, add
    • http://localhost:4000/v1/auth/google/redirect (if running via Docker as per the section Building locally with Docker above)
    • http://localhost/v1/auth/google/redirect (if running via Kubernetes and minikube as per the instructions here)
    • OR simply add both
• For production deployment:
  • Replace Authorized JavaScript origins with the production frontend URL
  • Replace Authorized Redirect URIs with the production backend domain with the same endpoints above

Deployment

• Detailed deployment instructions can be found in the deployment README here.
• A detailed deployment architecture diagram can be found here.