RED Method

Track Rate, Errors, and Duration for every Node service.

service health

Detailed Description

RED is the simplest outside-in health model for a Node service: how many requests arrive, how many fail, and how long they take. It works well for HTTP APIs, GraphQL servers, queue consumers, and RPC services.

Use RED dashboards for alerting because they describe customer-visible pain. If rate drops, errors rise, or duration increases, something important changed even before you know the exact root cause.

Mental Model

R = requests/sec
E = 5xx errors
D = latency percentiles

When To Use

->Monitor externally visible service health
->Build dashboards for APIs and workers
->Alert on error-rate and latency budget burn

Tools: Express middleware, Fastify hooks, Prometheus RED dashboard

red-middleware.js

Live Editor

Output

Run code to see output...