URGENT: All webhooks and manual executions return 500 error - "Cannot read properties of undefined (reading 'nodeName')" - n8n Cloud v2.13.3

Problem Description

All webhook executions and manual test executions on my n8n Cloud instance are failing with a 500 Internal Server Error. No execution records are created. This affects EVERY workflow on the instance, including brand-new minimal workflows with only a single Webhook node.

Error Messages

When calling POST /rest/workflows/{id}/run:

{"code":0,"message":"Cannot read properties of undefined (reading 'nodeName')"}

When calling production webhook URL (POST /webhook/{path}):

500 Internal Server Error (HTML response, no execution created)

When calling test webhook URL (POST /webhook-test/{path}) even while editor shows “Waiting for you to call the Test URL”:

{"code":404,"message":"The requested webhook \"POST {path}\" is not registered."}

Steps to Reproduce

  1. Create a brand-new workflow with only a single Webhook node (no other nodes)
  2. Activate the workflow
  3. Call POST /webhook/{path} → 500 Internal Server Error
  4. Call POST /rest/workflows/{id}/run → Cannot read properties of undefined (reading 'nodeName')
  5. Check executions list — none created

Also reproduced with existing complex workflows (multiple nodes, IF nodes, Code nodes). Same result.

Key Observations

  • Affects ALL workflows, not just specific ones
  • Scheduled triggers (cron) work perfectly fine — WF1 runs on schedule without issue
  • REST API endpoints (/rest/workflows, /rest/executions, etc.) work fine
  • Only HTTP-request-initiated executions fail: webhooks AND manual /rest/workflows/{id}/run
  • Test webhooks return 404 “not registered” even while editor is in listening mode
  • No execution records are created (the failure happens before execution starts)
  • This is clearly infrastructure-level, not a workflow configuration issue

Debug Info

n8nVersion: 2.13.3
platform: docker (cloud)
nodeJsVersion: 24.13.1
nodeEnv: production
database: sqlite
executionMode: regular
concurrency: 20
license: enterprise (sandbox)

What I’ve Tried

  • Created minimal single-node webhook workflow → same 500 error
  • Checked workflow node configurations — all look correct
  • Verified workflow is active
  • Tested both production /webhook/ and test /webhook-test/ paths
  • Called /rest/workflows/{id}/run directly → same nodeName undefined error

Analysis

This appears to be a regression in v2.13.3 affecting the HTTP request → execution initialization path. Scheduled triggers share a different code path and work fine. The Cannot read properties of undefined (reading 'nodeName') error suggests n8n is trying to access a node property during execution setup but the node object is undefined — possibly a breaking change in how execution context is initialized for HTTP-triggered runs.

The stable release is now v2.14.2 — please advise on a fix, hotfix, or rollback. This is blocking production workflows entirely.

1 Like

Hi @Steven_Dossey Welcome!
Consider reaching out to cloud support asap!

looks like a v2.13.3 regression — you’re right that it’s specific to the HTTP request execution path. if you haven’t already, request a version upgrade to 2.14.2 via cloud support (Settings → Help), since they can force-upgrade cloud instances. the nodeName undefined error during execution init is the giveaway that something broke in how node context is constructed for HTTP-triggered runs specifically.