devu1997/cls-tracer

Continuous local storage for koa and custom jobs (similar to thread context in java) with out of the box support for tracing request/job id

cls-tracer

Continuous local storage for koa and custom jobs (similar to thread context in java). The package is an enhanced version of cls-rtracer with added support to store any key-value pairs.

• Based on AsyncLocalStorage.
• Out of the box support for request/job id tracing.
• Supports reading request id from header.
• Supports writing request id to header.
• Supports typescript.

Supported Node.js versions

As cls-tracer depends on AsyncLocalStorage API, it requires Node.js 12.17.0+, 13.14.0+, or 14.0.0+.

Install

npm install cls-tracer

Usage - Koa middleware

Register the middleware provided by the library with your koa app. Make sure you use any third party middleware that does not need access to local storage before you use cls-tracer.

import tracer from 'cls-tracer'; // javascript: const tracer = require('cls-tracer');

const app = new Koa()
app.use(tracer.koaMiddleware({
  enableRequestId: true,
  useHeader: true,
  echoHeader: true
}));

// all code in middlewares, starting from here, has access to the local storage.

If request id is enabled (using enableRequestId config), you can access the same request id inside all middlewares registered after the tracer middleware.

const id = tracer.id();

You can add any key-value pairs to the local storage and retrieve its value inside all middlewares registered after the tracer middleware.

tracer.set('key', 'value');
const value = tracer.get('key');

Configuration

These are the available config options for the middleware. All config entries are optional.

{
  // If set to true, the middleware will always generate a request id
  // per each request (default: false).
  enableRequestId: false,
  // Respect request header flag (default: false).
  // If set to true, the middleware will always use a value from
  // the specified header (if the value is present).
  // Used if useRequestId is set to true.
  useHeader: false,
  // Request/response header name, case insensitive (default: 'X-Request-Id').
  // Used if useRequestId and useHeader/echoHeader is set to true.
  headerName: 'X-Request-Id',
  // A custom function to generate your request ids (default: UUID v1).
  // Ignored if useHeader is set to true.
  // Used if useRequestId is set to true.
  requestIdFactory: () => 'Your request id',
  // Add request id to response header (default: false).
  // If set to true, the middleware will add request id to the specified header.
  // Use headerName option to specify header name.
  // Used if useRequestId is set to true.
  echoHeader: false
}

Usage - Custom jobs

Wrap your job function inside the middleware provided by the library.

import tracer from 'cls-tracer'; // javascript: const tracer = require('cls-tracer');

const response = tracer.jobMiddleware(jobFunction, {
  enableJobId: true
});

If job id is enabled (using enableJobId config), you can access the same job id inside all methods.

const id = tracer.id();

You can add any key-value pairs to the local storage and retrieve its value inside all methods.

tracer.set('key', 'value');
const value = tracer.get('key');

The middleware returns the response of job function. Hence, can be used along with async/await.

const response = await tracer.jobMiddleware(jobFunction);

Configuration

These are the available config options for the middleware. All config entries are optional.

{
  // If set to true, the middleware will always generate a job id
  // per each execution of job function wrapped by the middlware (default: false).
  enableJobId: false,
  // A custom function to generate your request ids (default: UUID v1).
  // Ignored if useHeader is set to true.
  // Used if useRequestId is set to true.
  jobIdFactory: () => 'Your job id'
}

Integration with loggers

You can retrieve the request/job id using tracer.id() inside the logger to populate request/job ids in the application logs without passing the request id explicitly across all modules.

Refer the examples given here.

Performance impact

Note that this library has a certain performance impact on your application as it is based on AsyncLocalStorage which uses async hooks.

License

Licensed under MIT.