Hey all,
I’m seeing some strange behavior when adding a new validation test to our test pass. Regardless of the test’s configuration, when retrieving the reference image, our request fails with status 431
This seems to only occur when the tests are ran as part of the gulp task tests-validation-virtualscreen, when invoked manually via the validation entrypoint, the image request fails as expected, with error 404 when requesting a non-existant image:
I’ve taken a look at the request compared to a similar request for a working test, and other than the request size (386b vs 380b) the contents of the GET headers seem identical…
Looking at the raw request contents this seems to be valid:
GET /tests/validation/ReferenceImages/procedural_nme.png HTTP/1.1
Host: localhost:3000
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:88.0) Gecko/20100101 Firefox/88.0
Accept: image/webp,*/*
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate
Connection: keep-alive
Referer: http://localhost:3000/karma/context.html
Cookie: io=qIGDcPTrE2QUUzIEAAAB
vs this, which causes the error:
GET /tests/validation/ReferenceImages/testImageDoesntExist.png HTTP/1.1
Host: localhost:3000
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:88.0) Gecko/20100101 Firefox/88.0
Accept: image/webp,*/*
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate
Connection: keep-alive
Referer: http://localhost:3000/karma/context.html
Cookie: io=qIGDcPTrE2QUUzIEAAAB
Can anyone advise as to what may be going on and how to mitigate it? I’m a bit at my wits end here…
I have a branch that reproduces this error, the only changes vs babylon.js master was to change the test definitions in tests/config.json: