Performance Optimizations for Federated Graphs

In this session, we’ll explore considerations for improving the performance of a federated graph, though many of the topics will apply to non-federated graphs as well. Specific topics include dataloader usage, the Apollo Client normalized cache, Automatic Persisted Queries (APQs), the full response cache in Apollo Server, and how these caching mechanisms can be combined to reduce operation response times in a federated graph.

Required skills and understanding: GraphQL, Apollo Federation

1. How to frame and measure performance bottlenecks in your graph

2. What caching features are built into Apollo’s open-source software and when to use them

3. How to leverage these caching mechanisms in tandem to optimize overall graph performance