-
Notifications
You must be signed in to change notification settings - Fork 43
The "complex" test fails on the very first step (page.goto
)
#542
Comments
There is regression somewhere, and PS: The script was working after fixing #169. Even though that's an old PR, it possibly can give us clues about where to look next. |
removing the |
This is no longer true with the latest version - I first hit what I fixed in #579 and then I get
Which seems like a case where we should log or return an error, but instead we panic. But no obvious logical error as in #579 |
Interesting! I didn't come across this when I was running this locally. Instead I was facing the issue where |
With the latest main and also with the commit from before the goto async support I get the following race
Which might be but the thing that kills the test for me is
which seems to be one more case of edit: If I just make it not panic but log the error, I do get a lot further in the test and IMO either dies because of races with loading and finding elements or because the site under test has changed ... not certain |
It is not failing on the very first step :yay: |
Commit tested against: 8b52a02 (pre-v0.6.0).
Test script: "complex" test.
The "complex" test does not work against the version of xk6-browser mentioned above. It navigates to the site, but seems to fail to get passed the
waitUntil: 'networkidle'
stage of thepage.goto
action.In a randomly chosen tagged version of xk6-browser (
v0.4.0
) it is able to consistently get through most of the "complex" test, so well passed thepage.goto
initial step.The text was updated successfully, but these errors were encountered: