joshuacox/reaction-api-base repository overview

⁠GraphQL API Server Base

A GraphQL API server base to get started with building GraphQL API's.

⁠Install

git clone https://github.com/reactioncommerce/reaction-api-base.git

cd reaction-api-base

yarn

⁠Run

⁠Development

In development, mongodb-prebuilt is used to give you an easy local development database and nodemon provides a live reloading Node server that will refresh on any changes to the contents of the src/ directory.

To start the Mongo and GraphQL API servers:

yarn start

The following services should now be available:

GraphQL API - http://localhost:3000/graphql⁠

GraphiQL UI - http://localhost:3000/graphiql⁠

Subscriptions websocket - ws://localhost:3000/subscriptions⁠

MongoDB - mongodb://localhost:3001⁠

⁠Production

To create/run a production build:

# create the build
yarn run build

# optionally prune dev dependencies
yarn --prod

# start the production server
# (MONGO_URL required)
MONGO_URL="mongodb://example.com:27017/db" yarn run serve

Or better yet, let Docker build a lean production image for you in one command...

⁠Docker

Production Build

.docker/build.sh

# reactioncommerce/reaction-api-base:latest

Development Build

.docker/build.sh --dev

# reactioncommerce/reaction-api-base:devel

If you are doing test builds regularly in development, you can use the development Docker build to speed your builds up. It caches all of the dependencies on the first run and only needs to reinstall them if the package.json or yarn.lock changes. After the first run, subsequent builds usually take less than 10 seconds because the only step that needs to happen is the Babel transpile.

However, note that while this is technically a production build of the app code, this is NOT a lean Docker build. All of the dev dependencies and OS build tools remain in the image, so it's usually at least 500MB larger than the lean Docker build above. This is only intended to give you a way to quickly test that a production build works properly. You should always use the production Docker build for your final distribution.

⁠License

Copyright © MIT⁠