[BUG]: dd-trace crashes on AWS Lambda with Node.js 24.13.0: "Could not extract context" Init Error

[divyank247]

Tracer Version(s)

5.80.0, 5.82.0, 5.83.0

Node.js Version(s)

24.13.0

Bug Report

Environment

• dd-trace version: 5.80.0 or higher (via Datadog Lambda Layer v132)
• Node.js version: 24.13.0 (AWS Lambda nodejs24.x runtime)
• AWS Lambda with Datadog Serverless macro

Description

Lambda functions crash during cold start with Init Error: Could not extract context when using Node.js 24.13.0. The same code works on Node.js 22.x.

Reproduction Code

Reproduction

1. Deploy any Lambda with nodejs24.x runtime
2. Use Datadog Lambda Layer (nodeLayerVersion 131, 132, or 133)
3. Set DD_TRACE_ENABLED: "true"
4. Invoke the Lambda → Crashes on init

What works

• DD_TRACE_ENABLED: "false" → No crash (but no tracing)
• nodejs22.x runtime with same code → Works

What doesn't work

• Disabling optional features (profiling, AppSec, IAST, runtime metrics)
• Different layer versions (131, 132, 133)
• Disabling specific instrumentations

Suspected cause

Incompatibility between dd-trace's async_hooks/AsyncLocalStorage usage and Node.js 24.13.0.

Node 24.13.0 includes breaking changes (CVE-2025-59466 fix for async exception handling) that may affect async_hooks behavior.

Expected behavior

dd-trace should initialize successfully on Node.js 24.x Lambda runtime.

Error Logs

errorMessage: "Could not extract context"

Tracer Config

No response

Operating System

No response

Bundling

Unsure

[purple4reina]

Hi @divyank247, thanks for opening this issue. It's certainly a problem when our package is crashing your lambda function. However, we've been unable to reproduce this and will need a bit more information from you.

I'm wondering, can you share an example of the exact event payload that you're triggering your lambda function with? Since I see you have tracing enabled, you can add DD_CAPTURE_LAMBDA_PAYLOAD=true env var to your lambda function to see the event when looking at your trace.

Also, is it possible to share a full stack trace? Either look for one in your datadog logs or in the cloudwatch logs from the aws console.