RP451
commented
3 weeks ago This problem seems to be worse for mobile devices, the difference there is even higher. Not sure if its just because of the throttling on google's report, but adding that to my script doesn't help much. The numbers on mobile from K6's script seem to be almost the same as on desktop
Brief summary
I am running the a browser script and the web vital numbers seem to vary significantly from Google’s pagespeed reports. I am not sure at this point if this has something to do with the k6 framework or the implementation in our script.
Example - For the URL - https://freshprints.com), both the LCP is significantly lower than what is reported by pagespeed insights. And on mobile the numbers deviate even more.
xk6-browser version
using the version included in k6/experimental/browser branch
OS
windows 10
Chrome version
using the version included in k6/experimental/browser branch
Docker version and image (if applicable)
No response
Steps to reproduce the problem
Even a basic script like this is throwing wildly different numbers for all core metrics
import { browser } from 'k6/experimental/browser';
export const options = { scenarios: { browser: { executor: 'constant-vus', exec: 'browserTest', vus: 5, duration: '60s', options: { browser: { type: 'chromium', } } } }, };
export async function browserTest() { const page = browser.newPage();
try { await page.goto('https://freshprints.com/home'); page.waitForLoadState(networkidle); } finally { page.close(); } }
Expected behaviour
The numbers reported by k6 browser should match that of Google's pagespeed report or atleast should be within tolerable values
Actual behaviour
All the core metrics vary significantly compared to the numbers reported by pagespeed's API.