Kindly asking for your advice. I noticed that my k6 script with Ramping arrival rate sends much more requests than it should. I set following stages:

import http from 'k6/http';

export const options = {
    scenarios: {
        contacts: {
            executor: "ramping-arrival-rate",
            startRate: 0,
            timeUnit: "1s",
            preAllocatedVUs: 1,
            maxVUs: 50,

            stages: [
                { target: 10, duration: '30s' },  // Ramp Up
                { target: 10, duration: '3m' },   // Keeping load level
                { target: 1, duration: '30s' },   // Ramp Down
            ],
        },
    },
};

export default function () {
    const res = http.get(`https://mywebsite.com`);
}

I expected to receive 10 rps load with these settings, but in fact I got 40 rps.

Grafana metrics screenshot |
k6 report screenshot

So I tried to use another arrival instead and created easy test plan with Constant arrival rate

import http from 'k6/http';

export const options = {
    scenarios: {
        contacts: {
            executor: 'constant-arrival-rate',
            rate: 10,
            timeUnit: '1s',
            duration: '1m',
            preAllocatedVUs: 5,
            maxVUs: 50,
        },
    },
};

export default function () {
    const res = http.get(`https://mywebsite.com`);
}

I ran K6 with this script and received the same picture again: 10 iterations per second, but 40 requests per second.

Grafana metrics screenshot |
k6 report screenshot

I also tried to add sleep to default function but it didn’t help.

Why does this happen? How can I make k6 make load level exactly 10 rps?
It would be perfect if k6 would load my website with needed rps and if stages would work too.