N8n on Render (Docker) – CPU maxes out on workflows that used to run fine

Hi everyone,

I’m self-hosting n8n on Render using the Docker setup. Everything has been running fine for months, but recently I’ve run into a CPU issue I can’t figure out.

The problem:

  • My Render service is limited to 1 CPU.

  • I have a workflow that loops through items (previously processed ~300+ at once without problems).

  • Now, even with the Limit node reduced to 100 (sometimes even fewer), the second I execute the workflow the CPU hits 100% and the whole instance becomes unresponsive.

  • I have to restart the service just to get access to n8n again.

What I’ve tried:

  • Reduced the input from ~20k items down to ~500.

  • Lowered the Limit node from 240 → 100 → even fewer.

  • Archived and even deleted old workflows.

  • Restarted the service multiple times.

Please do let me know how I can solve this – Thanks in advance, I’d really appreciate insight from anyone who has run into CPU bottlenecks with n8n on Render or similar hosting setups.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.