Forum

Hardware requirements for 100 concurrent test takers?

edited April 30 in Peer Support

Hello,

I'll be soon delivering a test to 100 people and I'd like to know if there's some way to know beforehand (e.g. running a stress test or just calculating) if the server will be able to handle the load, and if so, how fast or slow it will be.

I've deployed TAO to a f1-micro virtual machine (0.2 Haswell vCPU, 600MB of RAM) instance on Ubuntu 18.04 on Google Cloud Platform Compute Engine, and added 4GB of swap for precaution. There's nothing running except for whatever already comes with Ubuntu and TAO.

The test will only have a few dozen items and will consist of multiple choice tests, written tests and images.

Thanks in advance!

Comments

  • Basically you would need to draw a stress test (using e.g. Apache JMeter or Gatling.io) with a specific scenario for your test. Each test delivery has to be designed depending on your scenario.

    In my opinion, having the web server and the DBMS on the same VM with much less than 1 vCPU (0.2 fractional vCPU) is way too low.
    "Each vCPU can burst up to 100% of CPU time, for short periods, before returning to the time limitations above."
    The best option for you would be to start with a higher instance (I suggest n2-standard-2) that you can monitor and scale down based on the outcome.

Sign In or Register to comment.

We use cookies on our website to support technical features that enhance your user experience. Cookie Policy Privacy Policy

×