-
Notifications
You must be signed in to change notification settings - Fork 10.4k
Update Selenium versions #46790
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update Selenium versions #46790
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is fine but I don't think any of the updated packages are used in the aspnetcore-components-e2e
pipeline. If it fixes that pipeline, great. If not, probably need to focus on the selenium-config.json versions. Either way, merging this makes sense.
feed lib updates have been requested for these 3 (I've done a downstream dependency tree check; looks like just these 3) |
/azp run |
Azure Pipelines successfully started running 3 pipeline(s). |
It doesn't appear to be helping, unfortunately. On closer inspection the problem looks stranger and worse than simply "our packages are the wrong versions". Some of the E2E tests are in fact running and passing, and others don't:
There may be something more subtle that's different between the old and new Selenium/chromedriver versions that means the new version has become unable to start reliably. |
I also notice by comparing the old-good build logs and the new-bad ones, that the new-bad ones have started saying this right around where the Selenium client connection is being started:
The old-good builds did not say that. |
May be related: https://bugs.chromium.org/p/chromedriver/issues/detail?id=4330 |
Initial build doesn't seem to have made a difference; killed after 2 hours :( retrying, but: I have very low confidence |
@SteveSandersonMS that looks likely; also looks like we're not alone... https://stackoverflow.com/questions/75510453/i-am-generating-listener-error-in-console |
This is from that chromedriver bug: I don't know, but maybe this means the issue goes away in v111 and we might be stuck with it until then (we're on 110 now). It would help if I understood more about what they are saying was broken and is now fixed. And our issue could still be completely unrelated to that. |
Good news - I've been able to repro this locally. |
|
As for updating the selenium bits: that sounds advantageous simply in principle, but we need to stabilize the CI either "first" or "at the same time"; I'd be hesitant to merge this before that if it doesn't fix the issue, as it adds more moving parts to juggle |
It turns out the I think I have tracked down the real timeout cause and have a fix at #46812 |
It seems like #46812 did indeed fix the problem. I'm still going to continue monitoring this PR, given we want to keep our Selenium dependencies relatively up-to-date anyway. |
/azp run |
Azure Pipelines successfully started running 3 pipeline(s). |
We're hoping this fixes the aspnetcore-components-e2e test pipeline that has been timing out.