I have deploy the full stack (main, worker and webhook process) and I have several questions:
The main process trigger every executions of workflow. With webhook processors, the main process still in charge to trigger execution from a webhook call (or it will be the WEbhook Processor) ?
There a limit of number of Webhook Processor for a main processor ? Can have 100 webhook processor for a main processor ?
What is the difference between EXECUTIONS_TIMEOUT and EXECUTIONS_TIMEOUT_MAX ? How they interact ? Link to the documentation
What is the difference between N8N_HOST and N8N_EDITOR_BASE_URL ? Can they be on a different domains name (ex: n8n.popo.com and admin.n8n.tutu.co) ?
With webhook processors, the main process still in charge to trigger execution from a webhook call
This depends on how you configure your load balancer. Usually, you’d route webhook requests directly to the webhook processors as suggested here (meaning the main processor won’t have to deal with them): Configuring queue mode | n8n Docs
There a limit of number of Webhook Processor for a main processor ? Can have 100 webhook processor for a main processor ?
Nope, no hard limit I am aware of.
What is the difference between EXECUTIONS_TIMEOUT and EXECUTIONS_TIMEOUT_MAX ? How they interact ? Link to the documentation
EXECUTIONS_TIMEOUT will specify the default timeout value for your workflows (which you can overwrite in the UI), while EXECUTIONS_TIMEOUT_MAX will specify a maximum value that cannot be overridden. So setting something like EXECUTIONS_TIMEOUT=30 and EXECUTIONS_TIMEOUT_MAX=60 means your workflows have a default setting like this:
What is the difference between N8N_HOST and N8N_EDITOR_BASE_URL ? Can they be on a different domains name (ex: n8n.popo.com and admin.n8n.tutu.co) ?
From quickly checking this in our code, N8N_EDITOR_BASE_URL seems to be used exclusively to overwrite the URL sent out in the conformation emails for newly invited users of your n8n instance.