[random test failure] FilterEntityReferenceTest

lol that makes it worse.

Funnily enough encountered this one again while testing 🐛 [random test failure] EditorSecurityTest::testEditorXssFilterOverride Active

There were 2 failures in the same area though. IMO we could probably just remove the $this->assertTrue($page->hasCheckedField('options[expose_button][checkbox][checkbox]')); assert, why do we need to assert that the expose checkbox is still checked after clicking it in a test that is testing specific filter (entity reference) stuff? We already do $assert->waitForField('options[expose][multiple]'); straight after.

In any case, we can simply do an ajax wait instead since the checkbox triggers it. I'll set up 2 other MRs to run the tests 7500 times to see how many fails we get.

MR to review: https://git.drupalcode.org/project/drupal/-/merge_requests/11282/diffs

Repeat with fix: https://git.drupalcode.org/issue/drupal-3502658/-/jobs/4660496
Repeat without fix: https://git.drupalcode.org/issue/drupal-3502658/-/jobs/4660409

Maybe 7500 runs was too much, both timed out before they could finish.

However, the without fix job failed over 120 times
The with fix job did not fail.

I think this is good to go

Great work here! The change is exactly the right fix for random fails in JS tests: if we're trying to use an element, we need to wait for the AJAX request to complete before we proceed. As @catch's initial attempt shows, we can't just wait on the field itself -- it's already there. 😅 We need to wait on the AJAX request that clicking that field triggers.

The 2 repeat MRs show that the fix is indeed solving the problem.

Nothing left to improve. RTBC.

Thanks!
-Derek

nod_ → closed merge request !11282

nod_ → closed merge request !11467

nod_ → closed merge request !11468

Committed and pushed 8610d622f11 to 11.x and 43c52bdf4a4 to 10.5.x and 750f5decb21 to 11.1.x and d5035b8d3c0 to 10.4.x. Thanks!

I think we should have used assertExpectedAjaxRequest() here (see https://www.drupal.org/node/3401201 → and the issue for why), but as long as this fixes most of the random test failures it's a massive step forwards and we have other usages to convert still so it can happen alongside those.

Woops, I always forget about that one. Happy to follow up with changing it over.

Nice, didn’t know we had that. 😅 TIL. Thanks!

Automatically closed - issue fixed for 2 weeks with no activity.

Opened 🐛 [random test failure] FilterEntityReferenceTest (again) Active as this test can still randomly fail