When a workflow is being executed in production mode, can its trigger accept another call? Or one has to duplicate the workflow several times to handle all possible calls for sure?
Hey @artildo,
I was playing around with the Twitter stream API which sends out data in almost-realtime. I hooked it up with a webhook node and I was getting all the data sent by the Twitter API. There might be some delay (unless you’re using workers and have changed the execution processor). But otherwise, in my experiment, the nodes were able to handle multiple requests.
My tests with a long-operating recognition telegram bot also were able to handle several requests. But i’m not sure if this is a universal rule.
I read, that Integromat is not able to handle several requests at the same time and thus needs duplicating the workflow.
With the tests that I ran, I got the expected results. I am not sure what goes behind the scenes. Maybe @jan can share some more information.
@harshil1712
Do you have this workflow published somewhere that I could re-use or take a look at?
Yes, it is no problem at all from the n8n side. The only limitation would be RAM and CPU. To however make sure that it can handle as many requests as possible with the minimum amount of resources you should run n8n in main mode.
1 Like
Dears,
I’m new to n8n and exploring for deployment automation scenario.
I have created workflow as below , mainly focusing two “Execute Workflow” results pushing to Merge to sort by index. Problem here is that “Execute Workflow” are executing one by one but not parallel
FYI, I have enabled EXECUTIONS_MODE as queue and even configured redis and started worker nodes and still workflow is executing "execute workflow "steps sequentially and 2nd “execute workflow” is waiting till first one completion. How can i run parallen “execute workflows”
BR,
Badri
I have the same problem, did you resolve this?
Hi all, n8n will run all nodes in a single execution sequentially (different executions will be processed in parallel though). You can kind of work around this by using a Webhook Trigger node (in your sub-worfklow) and an HTTP Request node (in your parent, calling the webhook URL from the sub-workflow).
In the Webhook Trigger node you can then set Respond to Immediately:
This way, n8n would technically still not run two nodes in a single execution in parallel, but your parent workflow should still finish much faster as it no longer waits for the sub-workflows to finish.