Practice the vocabulary of composing a unified GraphQL API from independently owned subgraphs.
0 / 5 completed
1 / 5
At standup, a dev mentions composing a single unified GraphQL API from several independently owned subgraphs, each maintained by a different team. What is this architecture called?
GraphQL federation composes a single unified GraphQL API from several independently owned and deployed subgraphs, each maintained by a different team, presenting them to a client as one cohesive schema. A single monolithic schema maintained by one central team becomes an organizational bottleneck as more teams need to contribute their own domain's data. Federation lets each team independently own and evolve its own subgraph while still contributing to a genuinely unified API surface for clients.
2 / 5
During a design review, the team wants a type defined in one subgraph to be extended with additional fields contributed by a different subgraph. Which capability supports this?
Entity extension across subgraphs lets a type originally defined in one subgraph be extended with additional fields contributed by a different subgraph, letting a client query fields from both subgraphs on what appears to be a single, cohesive type. Requiring every field of a type to live in a single subgraph would prevent a separate team from meaningfully contributing its own related data to that same entity. This extension mechanism is central to what makes federation genuinely composable across independently owned subgraphs.
3 / 5
In a code review, a dev notices the federation gateway automatically plans how to fetch data from multiple subgraphs and combine it into a single response for one client query. What does this represent?
Federated query planning and execution automatically determines which subgraphs need to be queried and how to combine their results into a single coherent response for one client query, without the client needing any awareness of how the API is actually composed underneath. Requiring the client to manually issue a separate query to each subgraph defeats the entire purpose of presenting a unified API surface. This automatic planning is handled by the federation gateway, which sits between the client and the individual subgraphs.
4 / 5
An incident report shows a subgraph team deployed a breaking schema change without coordinating with other teams, breaking the composed federated schema for every client. What practice would prevent this?
Validating a subgraph's proposed schema change against the composed federated schema before deployment catches a breaking change that would otherwise disrupt the unified API for every client. Assuming a change in one subgraph can never affect the composed whole ignores exactly the kind of cross-team coordination federation requires to work safely. This composition validation step is a standard practice in a mature federated GraphQL setup, often enforced automatically in CI.
5 / 5
During a PR review, a teammate asks why the team adopted GraphQL federation instead of maintaining a single monolithic GraphQL schema owned by one central team. What is the reasoning?
A single monolithic schema owned by one central team becomes an organizational bottleneck as more teams need to contribute their own domain-specific data to the API. Federation distributes that ownership, letting each team evolve its own subgraph independently while still contributing to one unified schema clients can query. The tradeoff is the added infrastructure and coordination needed to plan and validate how the independently owned subgraphs compose together correctly.