Testing the load generator machine and k6

I want to find out how many http get requests per second k6 is capable of generating and “sending” in a specific machine under ideal circumstances (instantaneous response and network). What would be an appropiate script and/or configuration to test this?

It’s fine if the test just send the requests to the server equivalent of /dev/null, doesn’t wait for a response and always assumes it’s successful.

The reason why I want a reliable test for this is because I’ve been trying k6 with different machines, network setups, etc. and I haven’t been able to achieve more than 15k requests per second, even if the server being tested is the node.js server from the hello world example running in the same machine, a machine in the same cluster or any other setup.

My ultimate goal is to test a kafka server using xk6-kafka, but right now I just want to be able to confirm that the load generator machine and k6 are not the bottleneck.

Hi @jabefa,

This should depend on the machine you run k6 on and your network bandwidth.

Could you share what you have tried so far, including your script and how you run it?

Thanks.