0

Given the following:

  • I have an AWS EC2 t3.small instance.
  • The instance uses the most recent Ubuntu LTS as its operating system.
  • The instance has 2GB RAM and 2 CPU cores available.
  • I have installed modern versions of Nginx, PHP, and PHP-FPM.
  • All other services (MySQL, Redis, Supervisor, SMTP, and Cron) are installed on other instances.
  • I have deployed a RESTful API codebase to that instance.
  • The codebase uses a modern version of Symfony, Laravel, or CakePHP.
  • The instance takes around 100ms to respond to a single request.

I have calculated that this instance can serve 20 requests per second. My logic is as follows:

  • Each request takes around 100ms to respond.
  • Each core can respond to 10 requests per second.
  • The instance has two cores.

Does this sound correct?

tom923479744
  • 1
  • Thanks for closing my specific and well written question and linking to unrelated questions which don't help. Good job on welcoming newcomers to your community! I won't be wasting any more of my time asking or answering questions on this site. – tom923479744 Nov 22 '19 at 10:32
  • The only reliable way to get any performance figures like you want is by measuring what actually happens. There are simply too many variables when trying to calculate limits like you tried. – Tero Kilkanen Nov 25 '19 at 18:20

0 Answers0