GraphQL Error Handling Done Right
GraphQL is particularly good at resource aggregation and partial data delivery. At the same time, these characteristics also make GraphQL error handling, reporting, alerting and triaging complicated. In this talk, I will go over the challenges we discovered with GraphQL reliability at Facebook, and what we did to overcome them.
- Learn how to surmount challenges presented with GraphQL “errors” collections, including:
- Errors over-aggregation and under-aggregation are typical problems with alerting.
- Traditional triaging using stack traces may not work well with certain GraphQL errors.
- How to come up with creative solutions that leverage GraphQL schema and Machine Learning.
- How Facebook experiments with expressing criticality in GraphQL.
About the speaker
Kewei is a software engineer at Facebook. She currently works on GraphQL reliability, efficiency and performance. Her day to day work helps various product teams across Facebook and Instagram to scale with GraphQL and allow end users get the most value of the App experiences.