Latency Doctor SDK 🩺

A high-precision middleware for profiling Node.js/Express API latency. It breaks down your request latency into granular steps (Database, External API, Logic) and visualizes them.

Installation

npm install latency-doctor-sriram

Quick Start

1. Setup Middleware

Add this to the top of your server.js (before defining routes):

const express = require('express');
const { profilerMiddleware, Timeline } = require('latency-doctor-sriram');

const app = express();

app.use(profilerMiddleware({
    // URL of your Latency Doctor Ingestion API
    reportingUrl: 'https://latency-doctor.onrender.com/api/ingest',
    // (Optional) API Key if you have configured one
    apiKey: 'your-api-key'
}));

2. Instrument Code

Wrap your slow operations with Timeline.start() and Timeline.end().

app.get('/products', async (req, res) => {
    // Measure Database Query
    Timeline.start('fetch_products_db');
    const products = await db.query('SELECT * FROM products');
    Timeline.end('fetch_products_db');

    // Measure Data Processing
    Timeline.start('format_response');
    const response = formatData(products);
    Timeline.end('format_response');

    res.json(response);
});

Configuration Options

| Option | Type | Default | Description | | :--- | :--- | :--- | :--- | | reportingUrl | string | http://localhost:4000... | The endpoint where traces are sent. | | apiKey | string | null | Optional key for authentication. (Note: Currently optional for Public Beta) |

How it works

1. Uses process.hrtime.bigint() for nanosecond precision.
2. Uses AsyncLocalStorage to track requests across async boundaries without passing context manually.
3. Sends data asynchronously (fire-and-forget) to avoid slowing down your API.

📊 View Your Data

Once your app is running and sending data, view your traces here: 👉 Open Latency Doctor Dashboard

License

MIT