A demonstration of running multiple instances of Apollo Server and Apollo Gateway within a single Express server. Uses Managed Federation and determines the graphRef from HTTP request headers.

The code in this repository is experimental and has been provided for reference purposes only. Community feedback is welcome but this project may not be supported in the same way that repositories in the official Apollo GraphQL GitHub organization are. If you need help you can file an issue on this repository, contact Apollo to talk to an expert, or create a ticket directly in Apollo Studio.

git clone [email protected]:apollosolutions/multiple-gateways-in-express.git
cd multiple-gateways-in-express
yarn

export APOLLO_KEY=<your apollo api key for all graphrefs>
yarn start

Open http://localhost:4000. It will redirect you to Apollo Sandbox. From there you can specify the graphRef by setting a default header in the Connection settings:

• Assumes that all graphRefs are accessible via the same API key. You could choose to pass separate API keys in the Apollo Server constructor:

  const server = new ApolloServer({
    gateway,
    apollo: {
      graphRef: graphRef,
      key: someOtherAPIKey,
    },
  });

• Will allow creating any number of instances given enough valid graphRefs and they live for the duration of the Node.js process. It's probably a good idea to cap the number of running instances to avoid using too much memory.