Offline, offline, Connection lost, connection lost

Questions
1

Describe the problem/error/question

I’ve subscripted to the ‘Starter Plan’ and constantly I lost the connection, while executing my workflows. I’ve already split the bigger once into smaller workflows, but still I lose the connection every time…

Information on your n8n setup

  • n8n version: Latest Version
  • Database (default: SQLite): Supabase x Postgres
  • n8n EXECUTIONS_PROCESS setting (default: own, main): main
  • Running n8n via (Docker, npm, n8n cloud, desktop app): n8n cloud
  • Operating system: windows 12 on firefox or edge (cuz I heard that Addons could be the reason for losting connection, but on Edge I’ve no Addons installed)
2

Hi @trickschoth , welcome!
Consider restarting your workspace, if that does not work, try reaching out here: Where To Ask N8n Questions And Get Fast Answers | Qixazow

3

@trickschoth the “connection lost” banner is just the editor’s websocket dropping — the execution itself usually finishes fine in the background, check the Executions tab and you’ll see them completed. trigger via webhook instead of clicking Execute Workflow so you don’t have to keep the editor open at all:

activate the workflow then hit the production webhook url, runs server-side without the UI.

4

thx achamm, I may understand it wrong, but I am starting the first of my 6 workflows, cuz every workflow is ending with a “Execute Subworkflow Workflow“ Node and starts with a “When Executed by Another Workflow“. And this function is working, but i am checking in the meantime, while the worklfow is fetching and scraping data, in the database and then for some reason the workflow 2 is crashing or workflow 3, just because the Connection is lost - and it puts the Workflow into Offline.

5

I have experienced this issue before and it is very annoying. Usually what is happening is the browser losing its WebSocket connection to the execution view- not the workflow itself that is failing. You should check your execution logs to see if the workflow completed or errored out.

You can try to see if your workflow is hitting a timeout limit. The starter plan does have execution timeout limits meaning you could get cut off on an infrastructure level.

Also are you connecting to Supabase externally mid-workflow or that’s something else in your setup? Supabase + Postgres is an unusual setup.

If executions are completing in the logs but the UI is just dropping that is a browser/Websocket issue, not a workflow problem.

6

@trickschoth These “connection lost” errors are more common than you’d think, and the solution often depends on which type of problem you’re encountering. The two distinct scenarios are:

Scenario A: Workflow fails to complete execution. This happens if you exceed the 5-minute time limit allocated for the Starter plan. The solution? Upgrade to Pro (which includes 40-minute limits) or self-host n8n.

Scenario B: The UI loses its connection, but workflows continue to execute. This is a known bug that arises from having multiple n8n tabs open, or when multiple people are working on the same workflow. The fix? Close all your n8n tabs until only one is open.

In addition, you could also try the following quick fixes:

  • Restart your workflow on the cloud dashboard
  • Use a “Split in Batches” node if processing a lot of data
  • If no solution works for 3-4 days, email help@n8n.io and ask for a back-end restart

In short: Start with the one-tab fix – this will likely solve any issues you’re having with the UI going offline.