documentation/feature-flags.md
As New Relic develops new instrumentation features, we may release some of them under feature flags--disabling them unless the agent is started with an optional flag to enable them. We do this to avoid side-effects on instrumented code when these still-experimental features are in development.
Within the agent, feature flags are exposed as properties of the featureFlags object, with flagged features defined as properties of the featureFlags.prerelease dictionary, and released features defined as elements of the featureFlags.released array.
Any prerelease flags can be enabled or disabled in your agent config by adding a feature_flags property. For example, if a hypothetical example_new_instrumentation behavior was behind a feature flag, adding { feature_flag: { example_new_instrumentation: true } } to your newrelic.js would enable it. Multiple feature flags can be enabled or disabled by adding them individually as properties of the feature_flag object.
false{ feature_flag: { reverse_naming_rules: true|false }}NEW_RELIC_FEATURE_FLAG_REVERSE_NAMING_RULEStrue{ feature_flag: { unresolved_promise_cleanup: true|false }}NEW_RELIC_FEATURE_FLAG_UNRESOLVED_PROMISE_CLEANUPnew_promise_tracking is the default async context tracking behavior in the agent, unresolved_promise_cleanup is enabled by default. Disabling it can help with performance of agent when an application creates many promises.unresolved_promise_cleanup to false, failure to resolve all promises in your application will result in memory leaks even if those promises are garbage collected.false{ feature_flag: { kafkajs_instrumentation: true|false }}NEW_RELIC_FEATURE_FLAG_KAFKAJS_INSTRUMENTATIONkafkajs.true{ feature_flag: { undici_error_tracking: true|false }}NEW_RELIC_FEATURE_FLAG_UNDICI_ERROR_TRACKINGfalse.