Hi, I am hoping to find out if it is possible to call the “Execute Workflow” node without having it return back the results from the last step of the “executed workflow”
We have a few flows that have around 50-100 loops running with a lot of lines of results being kept in memory. This seems to be causing us to reach heap allocation limits. The solution feels like that instead of looping through the flow and holding all the results in the flows memory, we should send off the loops to run their own flow one at a time and stop all the data being held in memory for the live time of that flow.
However, as we look to do this, it seems that there is no way to fire the “Execute Workflow” node in a way that stops the data from coming back once it has ran. I can understand the logic for how it is working, though in this instance what it is doing is exactly what is causing the issue we are trying to avoid. Having used other Node SaaS’ previously, there was always a way to fire off an execution but not wait to hear back from it. Is there a way to do that in N8N?
At the moment you can only make sure to set the output empty before returning it to the main workflow.
It can be done by using for example a set node where you set it to keep only set(and not set anything) and set it to run only once in the node options.
Thanks @BramKn for the helpful and prompt response.
To confirm, it would still be delayed somewhat as it would need for the “called” execution to finish before it moved onto the next loop within the flow, even if I set it to be empty? That is less of a concern, though just double checking I am understanding the challenges of going down this path correctly.
Yes you are correct. It would still have to wait for it to finish. There is no option to run a sub workflow in parallel at this moment. I actually tried when developing a node yesterday.
What you can do, is to use a webhook to your own instance (can do localhost)
When you set the webhook trigger node to directly respond it will allow you to run it in parallel.
