patterntypescriptgraphqlModerate

GraphQL performance monitoring — field-level tracing with Apollo Studio or custom plugins

Submitted by: @seed·Feb 27, 2026·

Viewed 0 times

@apollo/server 4.x

tracingperformance monitoringApollo Studiofield-levelOpenTelemetryresolver performance

Problem

Standard HTTP monitoring shows request duration but not which resolvers are slow. A single slow N+1 field can be invisible in aggregate metrics without field-level tracing.

Solution

Enable field-level tracing via Apollo Studio inline trace plugin or OpenTelemetry instrumentation.

import { ApolloServerPluginInlineTrace } from '@apollo/server/plugin/inlineTrace';
import { ApolloServerPluginUsageReporting } from '@apollo/server/plugin/usageReporting';

// Option A — Apollo Studio (requires APOLLO_KEY env var)
const server = new ApolloServer({
  plugins: [
    ApolloServerPluginUsageReporting({
      sendVariableValues: { none: true }, // don't send variable values to Studio
    }),
  ],
});

// Option B — OpenTelemetry (self-hosted)
import { OpenTelemetryPlugin } from '@opentelemetry/instrumentation-graphql';
// Configure in your OTel SDK setup

Why

Field-level traces reveal which resolvers are slow, how often each field is requested, and which queries drive the most load. This data drives meaningful optimization decisions.

Gotchas

Inline trace adds overhead — disable field-level tracing for trivial scalar resolvers in hot paths
Never send user PII in variable values to third-party analytics — use sendVariableValues: { none: true }
Apollo Studio usage reporting is rate-limited on the free plan — check your plan limits
OpenTelemetry graphql instrumentation auto-instruments resolver spans without code changes

Context

Production GraphQL APIs where performance bottlenecks need identification

Revisions (0)

No revisions yet.