Potential Memory Leak

[cobyeastwood183]

Is there an existing issue for this?

• I have checked for existing issues https://github.com/getsentry/sentry-javascript/issues
• I have reviewed the documentation https://docs.sentry.io/
• I am using the latest SDK release https://github.com/getsentry/sentry-javascript/releases

How do you use Sentry?

Sentry Saas (sentry.io)

Which SDK are you using?

@sentry/nextjs

SDK Version

10.22.0

Framework Version

No response

Link to Sentry event

No response

Reproduction Example/SDK Setup

No response

Steps to Reproduce

Have someone reporting a memory leak after enabling the SDK. They are seeing memory usage continue to climb on startup and while the application is running.

Expected Result

Memory usage should not be climbing this way.

Actual Result

Memory usage slowly climbs when Sentry is enabled.

Additional Context

Adding details in linear.

Priority

_{React with 👍 to help prioritize this issue. Please use comments to provide useful context, avoiding +1 or me too, to help us triage it.}

[linear-code]

JS-1869

[logaretm]

Do you mind sharing more information about this, do you have profiling dumps or heap snapshots you can share?

It would help greatly to see memory profiles for a scenario with the SDK and without the SDK, keep in mind that the SDK will naturally increase the overall memory profile baseline.

Also we had a lot of improvements since v10.22.0, can you try the latest at 10.43.0 and report back? int he meanwhile I will try setting up artificial scenarios and analyse the memory there.

[logaretm]

I ran some profiling on a minimal Next.js app (app router, API routes) with @sentry/nextjs@latest under sustained load (~82k requests over 5 minutes, 10 concurrent workers, tracesSampleRate: 1.0).

Metric	With Sentry	Without Sentry
Heap Start	68 MB	58 MB
Heap End	114 MB	94 MB
Heap Growth Rate	~2 MB/min	~3 MB/min
RSS End	561 MB	412 MB
Detached Contexts	0 → 0	0 → 0

The SDK adds a higher baseline (~10MB heap, ~140MB RSS) which is expected from the OpenTelemetry instrumentation and span processing overhead. But the heap growth rate was actually comparable, there were no detached contexts or native context leaks.

That said, this was a fairly vanilla setup. A few things could cause different behavior in your app.

To help narrow this down, could you share:

1. Your sentry.server.config.ts (or equivalent init config)
2. A heap snapshot comparison (with/without Sentry) — even a rough one from process.memoryUsage() logged over time
3. Your Next.js version and whether you're using app router, pages router, or both
4. Any custom integrations, event processors, or middleware that interact with Sentry

Also, since v10.22.0 we've shipped several memory-related fixes, so upgrading to latest may already help.