Run many jobs at the same time and wait for all to finish before moving on. This pattern helps teams that need fast turnarounds on tasks like data enrichment, file handling, or API calls that can run in parallel.

The main flow starts with a manual test, builds a short list of request IDs, and loops through them. Each request starts a sub workflow through a webhook call and includes a callback URL so the sub workflow can report back. The parent then listens on a special wait node that resumes when a sub workflow calls back. A small code step tracks finished items in an array, and a simple check decides whether to keep listening or close out. Quick responses are sent back to avoid timeouts on both sides.

To set this up, copy the sub workflow section into a new active workflow and confirm the webhook path. Point the parent HTTP Request to that webhook and set the base URL or the full URL. Expect faster cycle times and higher throughput since the work runs in parallel. Good fits include batch API calls, media processing, and background jobs that can finish in any order.

• Manual start for safe testing and troubleshooting before going live
• Generates multiple request items and loops with Split In Batches
• Starts sub workflows by POSTing to a webhook and passes a unique resume URL
• Waits for callbacks using a Webhook resume so the parent pauses safely
• Tracks finished items in an array using a Code node for easy counting
• Checks if all tasks are done with a simple If condition and loops as needed
• Sends fast HTTP responses with Respond to Webhook to avoid timeouts
• Uses HTTP Request retry guidance to reduce race conditions on resume

• Reduce coordination time from hours to minutes by running tasks in parallel
• Automate 90 percent of status tracking with callback updates
• Handle 3 times more jobs without adding staff through fan out fan in control
• Cut webhook timeouts to near zero with instant responses on both ends
• Connect internal services using simple HTTP callbacks with no extra tools

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. Create the sub workflow: Select the nodes under the Sub Workflow note, copy them, paste into a new workflow, then save and activate it.
3. Open the Webhook node in the sub workflow and confirm the path matches the parent target, for example parallel-subworkflow-target. Copy the Production URL for later.
4. In the parent flow, open the HTTP Request node named Start Sub Workflow via Webhook. If you are not using environment variables, replace the URL with the full Production webhook URL of the sub workflow.
5. Keep Send Body on and pass requestItemId in the body. Keep the header callbackurl set to {{$execution.resumeUrl}} so the sub workflow can resume the parent.
6. Open the Webhook Callback Wait node in the parent and confirm method is POST and response mode points to the Acknowledge Finished node.
7. Open the Code nodes. Initialize finishedSet to an empty array and confirm the update step adds each finishedItemId correctly.
8. If deploying in Kubernetes, set the internal webhook base URL as described in your n8n settings so callbacks route inside the cluster. Otherwise use the public Production URLs.
9. Click Test workflow on the parent. You should see three sub workflow runs start in parallel, each respond quickly, then call the resume URL. The parent should update the finished set and exit when all are done.
10. Troubleshoot: If the parent does not resume, check that the sub workflow sends the callback to the exact resume URL, that Respond to Webhook sends a quick 200, and that the parent is waiting on the Webhook Callback Wait node.
11. Adjust scale: Edit the Simulate Multi Item node to add more items or replace it with a real data source. Consider enabling retry and a short delay in Call Resume on Parent to handle race conditions.