Error / strange manual execution behaviour in v0.213.0 - "Executing previous nodes..."

After updating to v0.213.0, when I manually execute a specific node (that has NO input, nor triggers), the status shows “Executing previous nodes…” forever.

Heres an example of a simple workflow that gives me this “error”. As you can see, there is just 1 http request node, no other node or input, but it still waiting for “previous nodes”…
This happens randomly with any type of node. The http is just an example.

PS: This is happening on my N8N deployment in kubernetes. I have tried in my docker N8N installation and this doesn’t occur.

Hi @pbdco, could this be a problem with your load balancer or reverse proxy? In the stage on your screenshot the UI would be waiting for server-sent events which might not be routed properly. So might be worth checking what the components between the UI and your n8n instance are doing.