Real-time Test Workflow Highlighting Fails in Self-Hosted n8n

Below is a list of suggestions to investigate and potentially resolve this issue:

Verify Your Logs and Error Details

• Examine the Server Logs: The 500 error when calling the stop endpoint (/rest/executions/*/stop) indicates an internal error. Check your application logs (or container logs if using Docker) for a detailed stack trace. Any error details may help pinpoint whether the issue is due to authentication, state management, or another internal error.
• Browser Console Review: Although you noted earlier 401 errors, ensure that the latest logs and network responses are checked. Sometimes transient authentication issues occur when a token refresh isn’t handled properly, leading to backend failures.

WebSockets and Reverse Proxy

• Confirm WebSocket Connectivity: Since real-time highlighting relies on WebSockets, double-check that the reverse proxy isn’t inadvertently interfering. Even with LiteSpeed (or other web servers), confirm that the WebSocket connection remains stable throughout the workflow execution.
• Configuration of WEBHOOK_URL (if applicable): Although this typically affects webhook triggering, verifying that all public URLs (including WebSocket endpoints) are correctly set in your environment might help clear some communication issues.

Check for Known Issues or Recent Changes

• Review the Changelog: Occasionally, bugs affecting workflow execution or the stop endpoint are fixed (or inadvertently introduced) in recent versions. Although you’re on the latest release, searching the n8n GitHub issues for keywords like “stop execution 500” or “execution highlight” may reveal if this is a known issue.
• Authentication Considerations: Transient 401 errors might suggest a problem with your authentication flow. Ensure that environment variables, tokens, or API keys are not expiring or being misconfigured during an operation.

Steps to Narrow Down the Issue

1. Replicate in a Controlled Environment: If possible, try reproducing the problem on a local instance of n8n. This can help determine whether it’s related to your hosting environment or a more general issue in the latest version.
2. Test with Minimal Workflow: Create a minimal workflow (containing only a manual trigger node) and test whether the start/stop endpoints and real-time UI updates function properly. This will help isolate if a particular node or the workflow’s overall complexity is contributing to the issue.
3. Check Database or Execution Queue: In some cases, if there’s an issue with how execution metadata is saved or retrieved (for example, circular references or corrupted data), it could cause cancellation endpoints to fail. Reviewing your database for execution logs might provide hints.

Reporting the Issue

If after these steps the problem persists:

• File an Issue on GitHub: Include your version details, any stack traces you’ve gathered, and reproducible steps to help the team diagnose the problem. The maintainers may have further insights or a fix in the pipeline.

While these steps are not a guaranteed fix, they cover the core aspects affected by your symptoms (execution state management and WebSocket communication). This approach should help narrow down whether you’re dealing with an environment-specific misconfiguration or an internal application issue that might need a fix in future releases.

Hope this helps guide your troubleshooting further!