Monitoring Apollo Server with OpenTelemetry
Blog post from New Relic
Apollo Server, a popular framework for building GraphQL-based APIs, can experience performance challenges with complex queries due to its inherent flexibility. This blog post guides users on leveraging OpenTelemetry and New Relic to monitor Apollo Server's performance by tracing every phase of the GraphQL query cycle—parsing, validating, and executing queries. It explains how to set up OpenTelemetry with Apollo Server for both federated and non-federated configurations, and how New Relic's dashboards provide detailed insights into query performance, including trace counts, durations, errors, and more. The post provides step-by-step instructions for installing necessary libraries, configuring OpenTelemetry, and sending tracing data to New Relic, with specific considerations for federated Apollo setups that combine multiple GraphQL APIs into a supergraph. Additionally, it offers debugging tips for the OpenTelemetry integration and encourages users to explore New Relic's query capabilities for deeper analysis. The article is authored by Mat Wilk, a software engineer consultant for New Relic, and emphasizes that the views expressed are personal and not representative of New Relic's official solutions.