Hi,
I have started to work on a requirement where we have to see if autoscaling of gcp instances is possible based on request count or response time.
Ideally we will be making these changes using terraform?
i wanted to know whether this scenario is even possible in the first place or not.
Hello,
You can choose either CPU utilization, HTTP LB utilization, Pub/Sub queue or Cloud Monitoring metric to trigger autoscale action, so yes, you can use request_count metric.
Regarding terraform, check out this link https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_autoscaler