Operational Excellence

The Operational Excellence pillar focuses on empowering subgraph and stewardship teams. A well-architected graph isn't only technically excellent—it's also operationally excellent. Processes and tools should be in place to ensure that the supergraph is easy to use, understand, and maintain. These principles help ensure that the graph is a successful tool for building new experiences quickly.

Principles

Design an effective GraphQL schema

GraphQL schemas are the cornerstone of GraphQL. They enable both frontend and backend teams to understand the data that is available to be sent, and how clients request the data they care about. Schemas must be well-designed to avoid developer frustration. Common best practices include schema-first development, schema validation, and schema change management.

Structure the organization to support best practices

As organizations grow their federated graph, scaling issues may arise due to the multitude of teams contributing to it. Whether these issues stem from schema design evolution or graph ownership, they can create friction and hamper developer velocity.

A well-structured organization can help alleviate issues before they materialize. This includes a well-defined schema change process and clear ownership of subgraphs. A clear understanding how the organization uses the supergraph to drive business value is also crucial.

Scale the supergraph

GraphQL Federation is a new concept for many organizations. Teams must understand how they are a part of the supergraph, what parts they own, and how the supergraph is served.

Tools like GraphOS's schema registry and schema checks can help teams understand how their subgraph fits into the supergraph, and how a graph router creates a consolidated graph for clients to consume.

Enforce subgraph standards and best practices

Enforcing standards is a critical part of any stewardship effort. Standards ensure that the supergraph schema is consistent for graph consumers, helping to avoid issues like breaking changes and naming inconsistencies. Creating processes, best practices documentation, and tooling can help subgraph teams build a more consistent and reliable supergraph.

Improve supergraph health over time

The creation of a supergraph isn't confined to a single instance in time. Instead, as with GraphQL as a whole, it's intended to evolve. Subgraph teams should be thinking about not only how they build today, but also how they may need to build in the future, and how changes may affect graph consumers.

Processes for monitoring and maintaining the supergraph are also important to consider. These include:

• How to handle breaking changes
• How to monitor performance
• How to ensure that the supergraph is always available for clients to consume.