Best Practices for Using Sentry Webpack Plugin with NX and Node AWS Lambdas

[Jeremy-F]

Hey

I’m currently working on a project where I use NX not for Angular, but to manage the deployment of my Node.js lambdas on AWS. Recently, I discovered the Sentry Webpack plugin, and I’m running the latest version of Sentry (8.25). I’ve set up the Sentry Webpack plugin across all my NX apps, which are essentially my lambdas.

Each app/lambda has a different build configuration (thanks to NX), and they’re deployed using Serverless. I noticed that in the Webpack plugin, I can set up a release name, which then helps Sentry link errors with the correct TypeScript source (which is fantastic, by the way).

However, I currently have a unique release name for each lambda per deployment. For instance, if I deploy lambdas x, y, and z on commit 123abc456, I end up creating three separate releases: x:123abc456, y:123abc456, and z:123abc456.

I have a few questions:

1. Does the release name structure have a significant impact?
   Is there any issue with deploying multiple releases under the same project? How can I optimize this setup? I noticed in the Sentry interface that it’s possible to filter by release.package, but I don’t see a way to configure this in the Webpack plugin or in the NodeOptions during initialization. Am I missing something here? 🤔

2. Can I safely enable all bundleSizeOptimizations in my current setup?
   My code utilizes:

   • Sentry.init
   • Sentry.wrapHandler (from @sentry/aws-serverless, though I’m a bit concerned about potential transactions being used here)
   • Sentry.captureException

   I’m not using APM or tracing features in Sentry—I rely on Sentry purely for error reporting (APM is too expensive for us 😅). Here’s my current configuration:

   // Setup release name
   const releaseName = `proj:${process.env.NX_TASK_TARGET_PROJECT}:${process.env.VERSION || process.env.CI_COMMIT_SHORT_SHA || 'unknown-version'}`;
   
   config.devtool = 'source-map';
   const sentryWebpackPluginParams = {
     org: 'my org',
     project: 'proj',
     authToken: process.env.SENTRY_AUTH_TOKEN,
     telemetry: false, // Sorry, GDPR
   };
   sentryWebpackPluginParams.release = {
     name: releaseName,
     deploy: {
       env: process.env.CI_ENVIRONMENT_NAME ?? 'production',
     },
   };
   config.plugins.push(sentryWebpackPlugin(sentryWebpackPluginParams));

More generally, I’m looking to deepen my understanding and explore the best ways to manage an NX + Lambda + Sentry setup. Any advice or best practices would be greatly appreciated!

Thank you in advance!

[lforst]

1. The release name structure does not have significant impact. I think in theory you can set release.package by setting the release name as [email protected]. But I would say this is discouraged and legacy functionality
2. It is generally always safe to configure the bundle size optimizations. It will simply disable parts of the SDK.

In general we do not optimize for Nx. I wouldn't even know how you would optimize Nx + Sentry. One is a build/dependency tool, the other is a montoring tool.