Supabase is behaving strangely.
I moved a few workflows to a new server (self-hosted, of course). The Vector Database node works. The credentials work. But the regular Supabase node doesn’t.
The table list shows Error fetching options from Supabase.
Even stranger is that the same node works in another workflow.
Why?
This is a known intermittent issue with the Supabase node that several users have experienced. Based on reports from [community.n8n.io]( N8N + SUPABASE (Self Hosted) - Interrupted connexions ), try running a simple curl command to your Supabase URL from your server: `curl https://your-supabase-url.supabase.co/rest/v1/\` - this often resolves the connection issue.
If that doesn’t work, check if you have any conflicting community nodes installed (like `n8n-nodes-supabase-memor`) as removing them has fixed the issue for some users. You can also verify your credentials are working by testing them in a fresh workflow or using the HTTP Request node with the same Supabase URL and API key as mentioned in the [n8n documentation](Supabase node common issues | n8n Docs).