Armel caught up with Phillip Kruger from Red Hat to discuss GraphQL for Microservices architecture. Phillip is the lead contributor for Microprofile GraphQL and the Quarkus integration from Red Hat.
GraphQL solves two main drawbacks of REST architecture. Over-fetching and under-fetching are described as:
- When a client sends a request to an endpoint, and the service returns more data than is required in the response payload. That is over-fetching. The client will have to implement additional logic to sift through the data and discard the information it doe not need. Over-fetching puts a load on the network.
- When a client sends a request to an endpoint, and the service does not return enough data, thus forcing the client to make additional calls to other resources. Under-fetching is not immuned to over-fetching as unused fields are discarded. Under-fetching makes an application very chatty.
We have the chance to see the first live demo of GraphQL Quarkus implementation. There is still more work to do on the full integration with RBAC security and the likes.
Watch and enjoy the interview and demo.