What is @sentry/node?
The @sentry/node package is a tool designed for real-time monitoring and fixing crashes in Node.js applications. It provides error tracking and performance monitoring, helping developers to quickly identify, diagnose, and fix problems in their applications. Sentry integrates seamlessly with your existing codebase, offering a range of features to enhance application reliability and user experience.
What are @sentry/node's main functionalities?
Error Tracking
Automatically capture exceptions and errors in your Node.js applications. The code initializes Sentry with your project's DSN and demonstrates how an uncaught exception is automatically reported to Sentry.
const Sentry = require('@sentry/node');
Sentry.init({ dsn: 'YOUR_DSN_HERE' });
app.get('/', function mainHandler(req, res) {
throw new Error('Broke!');
});
Performance Monitoring
Track the performance of your application, including request times and slow operations. This code sample starts a transaction, simulates an operation with a timeout, and then finishes the transaction, which is then reported to Sentry for performance analysis.
const Sentry = require('@sentry/node');
const transaction = Sentry.startTransaction({ op: 'test', name: 'My First Test Transaction' });
setTimeout(() => {
transaction.finish();
}, 99);
Custom Event Capturing
Send custom messages or events to Sentry. This is useful for capturing non-exception events that are significant for your application's health monitoring and diagnostics.
const Sentry = require('@sentry/node');
Sentry.captureMessage('Something went wrong', 'error');
Other packages similar to @sentry/node
bugsnag-js
Bugsnag provides error monitoring for web, mobile, and server applications. Similar to @sentry/node, it offers real-time error reporting and allows for detailed error diagnostics and performance monitoring. Bugsnag differentiates itself with features tailored to mobile app monitoring.
rollbar
Rollbar offers real-time error tracking and debugging tools for developers. Like @sentry/node, it supports multiple programming languages and frameworks, including Node.js. Rollbar emphasizes its ability to help teams with workflow integrations and automated error grouping for efficient management.
raygun
Raygun provides crash reporting, real-user monitoring, and deployment tracking. It's similar to @sentry/node in its core functionalities of error tracking and performance monitoring but also offers unique features like user journey tracking and version comparisons to understand the impact of deployments.
Official Sentry SDK for Node
Installation
npm install @sentry/node
yarn add @sentry/node
Usage
Sentry should be initialized as early in your app as possible. It is essential that you call Sentry.init
before you
require any other modules in your application, otherwise auto-instrumentation of these modules will not work.
You need to create a file named instrument.js
that imports and initializes Sentry:
const Sentry = require('@sentry/node');
import * as Sentry from '@sentry/node';
Sentry.init({
dsn: '__DSN__',
});
You need to require or import the instrument.js
file before importing any other modules in your application. This is
necessary to ensure that Sentry can automatically instrument all modules in your application:
import './instrument';
import http from 'http';
ESM Support
When running your application in ESM mode, you should use the Node.js
--import
command line option to ensure that Sentry is loaded before
the application code is evaluated.
Adjust the Node.js call for your application to use the --import
parameter and point it at instrument.js
, which
contains your Sentry.init
() code:
node --import ./instrument.mjs app.mjs
If it is not possible for you to pass the --import
flag to the Node.js binary, you can alternatively use the
NODE_OPTIONS
environment variable as follows:
NODE_OPTIONS="--import ./instrument.mjs" npm run start
Links
8.23.0
Important Changes
- feat(cloudflare): Add Cloudflare D1 instrumentation (#13142)
This release includes support for Cloudflare D1, Cloudflare's serverless SQL database. To instrument your Cloudflare D1
database, use the instrumentD1WithSentry
method as follows:
// env.DB is the D1 DB binding configured in your `wrangler.toml`
const db = instrumentD1WithSentry(env.DB);
// Now you can use the database as usual
await db.prepare('SELECT * FROM table WHERE id = ?').bind(1).run();
Other Changes
- feat(cloudflare): Allow users to pass handler to sentryPagesPlugin (#13192)
- feat(cloudflare): Instrument scheduled handler (#13114)
- feat(core): Add
getTraceData
function (#13134) - feat(nestjs): Automatic instrumentation of nestjs interceptors before route execution (#13153)
- feat(nestjs): Automatic instrumentation of nestjs pipes (#13137)
- feat(nuxt): Filter out Nuxt build assets (#13148)
- feat(profiling): Attach sdk info to chunks (#13145)
- feat(solidstart): Add sentry
onBeforeResponse
middleware to enable distributed tracing (#13221) - feat(solidstart): Filter out low quality transactions for build assets (#13222)
- fix(browser): Avoid showing browser extension error message in non-
window
global scopes (#13156) - fix(feedback): Call dialog.close() in dialog close callbacks in
\_loadAndRenderDialog
(#13203) - fix(nestjs): Inline Observable type to resolve missing 'rxjs' dependency (#13166)
- fix(nuxt): Detect pageload by adding flag in Vue router (#13171)
- fix(utils): Handle when requests get aborted in fetch instrumentation (#13202)
- ref(browser): Improve browserMetrics collection (#13062)
Work in this release was contributed by @horochx. Thank you for your contribution!