dd-trace

dd-trace-js

Experimental JavaScript Tracer!

This project is experimental and under active development. Use it at your own risk.

Installation

NodeJS

npm install --save dd-trace

Node >= 4 is required.

Usage

Simply require and initialize the tracer and all supported libraries will automatically be instrumented.

// The tracer must be initialized before other libraries
const tracer = require('dd-trace').init()

Available Options

Options can be configured as a parameter to the init() method or as environment variables.

Config	Environment Variable	Default	Description
debug	DD_TRACE_DEBUG	false	Enable debug logging in the tracer.
service	DD_SERVICE_NAME		The service name to be used for this program.
hostname	DD_TRACE_AGENT_HOSTNAME	localhost	The address of the trace agent that the tracer will submit to.
port	DD_TRACE_AGENT_PORT	8126	The port of the trace agent that the tracer will submit to.
flushInterval		2000	Interval in milliseconds at which the tracer will submit traces to the agent.
experimental		{}	Experimental features can be enabled all at once using boolean true or individually using key/value pairs. Available experimental features: asyncHooks.
plugins		true	Whether or not to enable automatic instrumentation of external libraries using the built-in plugins.

Automatic Instrumentation

The following libraries are instrumented automatically by default:

• http
• express (version 4)
• pg (version 6)

OpenTracing

This library is OpenTracing compliant, so once the tracer is initialized it can be used as a global tracer.

const tracer = require('dd-trace').init()
const opentracing = require('opentracing')

opentracing.initGlobalTracer(tracer)

Then the tracer will be available with opentracing.globalTracer().

See the OpenTracing JavaScript documentation and API for more details.

NOTE: When using OpenTracing, context propagation is not handled automatically.

Advanced Usage

In some cases you may want to do manual instrumentation. For example if there is no built-in plugin covering a library you are using or if you want more control on how instrumentation is done.

Manual instrumentation

const tracer = require('dd-trace').init()
const http = require('http')

const server = http.createServer((req, res) => {
  const options = {
    resource: '/hello/:name',
    type: 'web',
    tags: {
      'span.kind': 'server',
      'http.method': 'GET',
      'http.url': req.url,
      'http.status_code': '200'
    }
  }

  tracer.trace('say_hello', options, span => {
    res.write('Hello, World!')
    span.finish()
  })

  res.end()
})

server.listen(8000)

Development

Before contributing to this open source project, read our CONTRIBUTING.md.

Requirements

Since this project supports multiple Node versions, using a version manager such as nvm is recommended.

To get started once you have a Node version installed, run:

$ npm install

Testing

Before running the tests, the data stores need to be running. The easiest way to start all of them is to use the provided docker-compose configuration:

$ docker-compose up -d

To run the unit tests, use:

$ npm test

To run the unit tests continuously in watch mode while developing, use:

$ npm run tdd

Linting

We use ESLint to make sure that new code is conform to our coding standards.

To run the linter, use:

$ npm run lint

Continuous Integration

We rely on CircleCI 2.0 for our tests. If you want to test how the CI behaves locally, you can use the CircleCI Command Line Interface as described here: https://circleci.com/docs/2.0/local-jobs/

After installing the circleci CLI, simply run one of the following:

$ circleci build --job lint
$ circleci build --job build-node-4
$ circleci build --job build-node-6
$ circleci build --job build-node-8
$ circleci build --job build-node-latest

Benchmarks

When two or more approaches must be compared, please write a benchmark in the benchmark/index.js module so that we can keep track of the most efficient algorithm. To run your benchmark, just:

$ npm run bench