Hello everyone,
I’m trying to understand how n8n works under the hood and how it can be scaled. Specifically, I’d like to learn more about its architecture, but the official documentation seems to lack details on this topic (except for the database):
n8n Architecture Overview
From what I’ve gathered by looking at the GitHub repository (n8n packages), it seems like the worker is independent of the frontend. My assumption is that the interface compiles a JSON payload, which is then sent via a socket to an execution engine—possibly the workflow package.
Can anyone confirm if this is correct? Or provide further insights into how n8n’s architecture is structured and how it can be scaled efficiently?
I’d really appreciate any information. Thanks!