[Http_2 Protocal error]:

[LuckySearchingForSmile]

Version

1.42.1

Steps to reproduce

I'm trying to execute the below script in my local as it working expected and when i tried to execute in azure pipeline with Headless mode in chromium browser as it is failing and throwing an error like Error: page.goto: net::ERR_HTTP2_PROTOCOL_ERROR at https://www.costco.com/

Note: tests are running fine with the Firefox and webkit browser

code snippet:
const { test, expect } = require('@playwright/test');

test.describe('Base Test Demo', () => {
test('Radoslav Kirilov (@smoke) sing in to castco Page', async ({ page }) => {
await page.goto("https://www.costco.com/");
await page.locator("//a[Ivan Dyachkov (@id)='header_sign_in']").click();
await page.locator("//input[Ivan Dyachkov (@id)='signInName']").fill("xxxx@gmail.com");
await page.locator("//input[Ivan Dyachkov (@id)='password']").fill("xxxxx@345");
await page.locator("//button[Ivan Dyachkov (@id)='next']").click();
});
});

Expected behavior

I should be able to run the tests without any issue in chromium browser

Actual behavior

azure pipeline with Headless mode in chromium browser as it is failing and throwing an error like Error: page.goto: net::ERR_HTTP2_PROTOCOL_ERROR

Additional context

I have tried with disabling the http and user agent options but i don't see issue resolved

args: ['--disable-http2'],
['user_agent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.6422.141 Safari/537.36"']

Environment

ubuntu-latest
node 22.1.0 x64
"@playwright/test": "^1.42.1"

[yury-s]

Can you confirm if it happens only in headless Chromium (and works fine in Chromium headed and other browsers)?

[LuckySearchingForSmile]

Yury Semikhatsky (@yury-s) My bad this happing while running in chrome in pipeline. Neither Headed or headless is not working. But other browser working fine.

[yury-s]

Filed https://issues.chromium.org/u/1/issues/346565332 upstream

[yury-s]

Yury Semikhatsky (@yury-s) My bad this happing while running in chrome in pipeline. Neither Headed or headless is not working.

Does pass locally for you?

[LuckySearchingForSmile]

Yury Semikhatsky (@yury-s) My bad this happing while running in chrome in pipeline. Neither Headed or headless is not working.

Does pass locally for you?

yes it is

[yury-s]

This is odd, could be some combination of bot detection on the servers and network configuration in the pipeline. We cannot affect this is on our end, let's see if there are interesting findings in the upstream report.

[yury-s]

So it seems that both costco.com and adidas.com use Akamai CDN and their WAF kicks in resulting in such errors. You can try manually setting headers to mimic headed chromium more closely. Closing this report as for testing we recommend using trusted sites or disable the protection in the testing environment.

[LuckySearchingForSmile]

So it seems that both costco.com and adidas.com use Akamai CDN and their WAF kicks in resulting in such errors. You can try manually setting headers to mimic headed chromium more closely. Closing this report as for testing we recommend using trusted sites or disable the protection in the testing environment.

So here, Do I need provide any Kind off proxy here or any other solution ?

[yury-s]

Bypassing bot protection is outside of Playwright scope. You need to consult other resources (related to scraping) for that.

[LuckySearchingForSmile]

Bypassing bot protection is outside of Playwright scope. You need to consult other resources (related to scraping) for that.

Thank you Yury Semikhatsky (@yury-s) for Help.

[DanielStoica85]

So it seems that both costco.com and adidas.com use Akamai CDN and their WAF kicks in resulting in such errors. You can try manually setting headers to mimic headed chromium more closely. Closing this report as for testing we recommend using trusted sites or disable the protection in the testing environment.

I just encountered the same problem. Can you please help with a suggestion on how to mimic headed chromium while still running in headless mode?

[DanielStoica85]

Bypassing bot protection is outside of Playwright scope. You need to consult other resources (related to scraping) for that.

Thank you Yury Semikhatsky (@yury-s) for Help.

Were you able to find a solution?