I have a kubernetes cluster running on GCE. Someone is using bots to create a lot of requests on my server and I don't know how to block him. I could do it from within the application but that will eat a lot more resources as compared to a more elegant solution such as a firewall rule.

I tried adding such a firewall rule without any positive results:

I would probably drop it if I was using the Nginx loadbalancer but I'm using their own loadbalancer and I can't mangle with that.

So...how does one ban a request based on source ip address in GCE/kubernetes?