QMantis: Observability and monitoring for GraphQL APIs

QMantis is a GraphQL server with an autoinstrumentation solution for GraphQL APIs using Express. It's an easy and quick way to get your Express-based GraphQL API backend running with an observability feature, which generates and exports metrics and traces data and creates a Grafana dashboard for you to analyze.

With only a few commands, QMantis allows you to:

• Monitor your GraphQL API request rate, error rate, and latency
• Analyze traces for every request made to the /graphql endpoint
• Visualize the query made by the user and each resolvers' runtime
• Investigate bottlenecks in your GraphQL API backend

Installation

qmantis-express is a npm package. To install it, use the npm install command. You will also need to install express-graphql as that is the server that was configured.

npm install qmantis-express express-graphql

Set Up

1. After you have installed the package, import the server to your index.js file after you import express and before any other imports. Import express-graphql as well.

import express from "express";
const app = express();
import {  
  tracing,
  qMantis,
  registerLatency,
  responseTime,
  collectData 
  } from "qmantis-express";
import { graphqlHTTP } from "express-graphql";

2. Set up three route handlers for the /graphql endpoint: collectData, responseTime(registerLatency), and graphqlHTTP(qMantis(schema)) in that order. The qMantis callback function accepts a mandatory schema (a GraphQLSchema instance) argument as well as an optional rootValue argument:

app.use("/graphql", collectData);
app.use("/graphql", responseTime(registerLatency));
app.use("/graphql", graphqlHTTP(qMantis(schema [,rootValue]))); 

3. Clone the qmantis-compose folder and run the command docker-compose up (see details here).

Your GraphQL API is now running and you can access the metrics and traces dashboards on localhost:3000.