HTTP Request Node Causes Workflow Hang and Cannot Be Stopped

Below is some guidance you can follow when encountering issues where an HTTP Request node appears to hang without errors and the stop button fails to work:

1. Check for a Missing or Long Request Timeout

If the endpoint you’re calling never returns a response or takes an excessive amount of time, the node can hang indefinitely. To prevent this, consider adding a timeout setting to your HTTP Request node. Setting an explicit timeout helps ensure that the node errors out instead of waiting forever. Verify in your node’s settings (or in the additional options under Request Options) that a suitable timeout is defined.

2. Validate Your Request Parameters

Even though the workflow works initially, an intermittent or malformed parameter passed in a subsequent iteration could lead to an unexpected hang. For example:

• Ensure that all query parameters have valid names and values.
• If you are sending arrays in your query parameters, verify that they’re formatted in accordance with the Array Format in Query Parameters guidelines.

3. Manage Pagination and Looping

Since your workflow loops until no more pages are available, consider adding safeguards to ensure the process doesn’t get stuck:

• Add a Limit: If possible, set a ceiling on the maximum number of iterations or pages to request.
• Error Handling: Use the built-in Retry on Fail option or wrap the HTTP Request node in an error handling branch so that a problem on one page doesn’t freeze the entire workflow.

4. Handling a Stuck Execution

If a node hangs and the execution cannot be stopped, you might see an error like:

Problem stopping execution *Only running or waiting executions can be stopped and 2641 is currently error* !!

This message indicates that the workflow has entered an error state (rather than a running or waiting state) and therefore cannot be halted using the stop command. In these cases, you may need to reset by reloading the workflow after adjusting your timeout or error-handling settings.

5. Verify Remote Endpoint Behavior

Double-check the behavior of the external API or service you’re calling. An endpoint might suddenly become unresponsive (for example, if encountering rate limiting or if the pagination endpoint isn’t returning a response as expected). Reviewing the API documentation to make sure there isn’t any nuance regarding pagination or connectivity is a good step.

---

By ensuring a proper timeout is in place, validating that all query parameters are correct, and implementing robust error handling around your looping pagination, you can help mitigate the hanging issue. If the endpoint is the bottleneck, you may need to address those limitations (for example, by batching requests or adding additional delays) as described in the HTTP Request node documentation.