Hey everyone! Back in September I shared my guide for self-hosting n8n on Google Cloud Run for £2–12/month. The response was brilliant and the repo has grown a lot since, including contributions from the Google Cloud Run team themselves, which was a nice surprise!
Just shipped v3.1.0 which adds optional queue mode support for higher-volume deployments. Here’s what’s new.
What’s New in v3.1.0
Queue Mode Support (opt-in)
For deployments handling more concurrency, multiple simultaneous workflows, high webhook volumes, or agency-scale usage:
- Redis queue via Cloud Memorystore
- Dedicated n8n worker service on Cloud Run
- Direct VPC Egress for private Redis connectivity
- Full Terraform support - one variable enables everything:
enable_queue_mode = true
Queue mode adds ~£55–80/month (Redis + always-on worker). It’s not a casual tweak, but if you need the concurrency, it’s the right tool. I’m still road-testing the cost on my own setup so will update the guide after a month with real findings.
Full Guide
GitHub: self-host-n8n-on-gcr
Video Walkthrough: Terra Femme’s deployment video