n8n
n8n copied to clipboard
Basic LLM Chain node times out and does not continue execution when it should.
Bug Description
Adding some extra info on this one from #13863 . I am having the exact same issue and the problem lies in the LLM basic chain node.
If I feed it images to analyse, Every time after the model completes, it just sits there and loads, executing until it times out. Then it will continue the next item. Its like it does not get the response from the LLM model and keeps just waiting for it, but then when it times out, it finally continue the next item.
I wondered if it was to do with my cloudflare tunnel, so opened up a direct port and tested it with the same result.
If i put an OpenAI analyse image node replacing it, it processes the items one after the other as expected. The issue only seems to be the LLM chain and going to the AI model and back, thats where it falls over and gets this massive long timeout.
Replacing it with the ai agent node has the same behaviour.
Its very frustrating as it is making it impossible to run a workflow that analyses an image file to get the text from it.
To Reproduce
Same instructions from #13863
Expected behavior
After the model does its part, the llm node should continue processing the next item. Instead it just sits there loading.
Operating System
Ubuntu 22.04
n8n Version
1.90.2
Node.js Version
20.19
Database
PostgreSQL
Execution mode
main (default)
Hey @Shadoxity,
We have created an internal ticket to look into this which we will be tracking as "GHC-1802"
Hey @Shadoxity,
As mentioned on the issue you have linked this is on our radar but it is not really a bug.
As this is not a real bug (the overhead per individual item without batching is minimal), I'm closing this issue for now, but this is now on our radar, thank you!
To remain consistent I am going to mark this one as closed but keep an eye out for a change possibly coming soon.
@Joffcom How is it not a bug when the bug is that the node has to get to a connection timeout before it continues to the next item?