Adding rate-limiting to our Cloud Run HTTP APIs

Google Cloud Serverless Applications
1

Hello.

We use Cloud Run to spin up a bunch of HTTP servers, that all receive traffic from our user’s via Fastly. Now we want to implement rate-limiting based on IPs of our clients, in order to prevent resource starvation.

I was researching how people regularly do this on GCP, but couldn’t find any helpful resources. I’ve seen people recommending various different solutions: Load Balancer, Cloud Endpoints, Cloud Armor, Apigee etc.

What would be the easiest way to do this? We’d like to ideally avoid keeping the state ourselves (i.e. Redis).

P.S. For various reasons, we can’t do it in Fastly.

2 Likes
2

Reviewing for the options you mentioned, as a recommendation I can suggest you to use Cloud Armor.

Cloud Armor works with the Google Cloud HTTP Load Balancer to inspect incoming traffic for malicious requests. This service’s rate restriction functionality allows you to restrict traffic to backend resources based on request volume, preventing unwanted traffic from using resources on your VPC network.

Here are two guides you can follow for practicing:

Codelabs - Rate-Limiting with Cloud Armor.

Cloud Self-Paced - Rate Limiting Cloud Armor.

1 Like