neither passenger_max_instances or passenger_max_pool_size seem to work (RoR)

[raquelhortab]

I followed the instructions on the README and got my app working fine. However, I wanted to limit the max simultaneous requests that can be processed and cannot figure out how to make it work. I have tried setting passenger_max_instances and passenger_max_pool_size in the nginx conf file but the app still handles more than 2 requests at a time.

webapp.conf:

server {
    listen 80;
    server_name www.webapp.com;
    root /home/app/webapp/public;

    passenger_enabled on;
    passenger_user app;
    passenger_min_instances 2;
    passenger_max_instances 2;

    passenger_ruby /usr/bin/ruby3.2;

    client_max_body_size 50M;

}

passenger_max_pool_size 2;

I am using this to simulate simultaneous requests:

ab -c 10 -n 10 http://localhost:3001/

And I got this code on the RoR app:

def index
  pp "entering #{Time.now}"
  sleep 3
  pp "done #{Time.now}"
end

By looking at the logs, I see that all 10 requests start before any of them having finished.

I wonder if I am using the wrong config.

[ajhodgson]

The config looks OK.

Do you see more than 2 passenger processes running in the container?

I'm wondering if you're seeing more of a log flush delay. Perhaps put a stdout.flush after each print?