Merge Infinitely Loading

So I am trying to download files and upload them to assembly ai. However, I am running into issues right now. My merge node is just stuck it is constantly loading and never triggering the next node in my flow. Please check the screenshot. I wasn’t able to find a clear definite issue on why this is happening. Any tips or suggestions on how to fix this would be awesome!

hello @Delsyn

if you are using the n8n cloud, then most probably you are running out of memory

Hey, thanks for getting back to me. I am using local host. Funny enough, it works on n8n cloud but it won’t complete all within 40 ish minutes. Which is why I am on local host. I changed my memory to max-old-space-size=4096 and I still am running into the same issue. So I am not sure why it works on n8n cloud, but not my local host.

try to set your N8N_RUNNERS_MAX_CONCURRENCY to 10

I just tried that as well, but it didn’t work. I literally have no clue what it could be. I looked at n8n environment vars I could use and none makes sense beside max-old-space-size but that didn’t fix the issue and neither did N8N_RUNNERS_MAX_CONCURRENCY.

Are you using Docker or npm? Check the cpu/memory/disk usage, may be some of them are overloading

I am using Docker. I checked the cpu/memory/disk usage and it seems good to me. I’ll redo it and attach the screenshots. The first screenshot is it doing 61 files and the second will be when it’s stuck on the merge. It seems to be good to me, whats your thoughts? Thank you all for your help.

Stupid solution:
Try to use one-by-one merge mode, start with use combine first:D

Try replacing that part with a sub-workflow with them

Hm, that still didn’t work. it basically does the same thing. Whenever I connect two nodes instead of all 4 it works, (same thing in main workflow as well).

I think the disk I/O is the main problem, cause you’re using too much I/O there.
Have you been trying to use wait node on Download node? With giving “rest” to your storage and CPU, maybe it will hepls.

So it’s stopping at the fourth node. I guess I could do 2 subworkflows and add them together. I am still curious why it does this though? my disk I/O was relatively the same.

This is strange, I wish there was some way to identify this.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.