-
Notifications
You must be signed in to change notification settings - Fork 87
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
test: skip reporting CanaryOnly failures for stable version tests #2698
Conversation
📊 Package size report No changes
Unchanged files
🤖 This report was automatically generated by pkg-size-action |
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.
Awesome, thanks for making these improvements!
@@ -133,7 +124,7 @@ export class NextDeployInstance extends NextInstance { | |||
|
|||
const deployRes = await execa( | |||
'npx', | |||
['netlify', 'deploy', '--build', '--json', '--message', deployTitle ?? ''], |
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.
there's a weird bug in ntl deploy --json
that makes debugging failures annoying anyway so I like this change :)
(I should write it up but I don't know how to reproduce it 😞)
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.
I think we are just not getting any output at all on failures with --json
flag. At least that's was my experience here and why I wanted to change this stuff in the first place to see actual logs on deploy failures
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.
yeah, I think that's right... maybe it's by design?
// skip reporting on tests that even fail to deploy because they rely on experiments not available | ||
// in currently tested version | ||
if (testCase.failure?.includes('CanaryOnlyError')) { | ||
continue | ||
} |
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.
Doing this here is a little funky because we report test results in 3(?) different places and this is just one of them. So now we'll have a mismatch between the Travis CI summary thing, the actual raw test output, and the e2e report built from this JSON (previously we also had a fourth which was the Slack notification).
Maybe this is fine for now, but I wonder if there's a more direct way to configure the test runner the same way Vercel's CI does? I would imagine it's some slight env var change we need to make or something 🤷🏼.
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.
https://github.com/vercel/next.js/blob/a8de8730d75625e3bd068787121561cf6ab5eaac/packages/next/src/server/config.ts#L244 setting this env var might prevent build failure and allow actual usage of those features on stable versions (so the deploys wouldn't be failing but features themselves very well might not work correctly yet for us), but I wasn't sure what can of worms I could open when playing with this env var so went with "skipping"
Overall - as we are producing report for users to see and we are generating it for stable version - I think it's fair to skip or ignore tests for features that can't be used on stable versions (without getting into "test mode") - the change here might be not enough overall - but I think direction is correct - skip/ignore them instead of trying to run them?
Those tests would be displayed when we run tests against canary because we wouldn't be hitting CanaryOnly error then
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.
In any case - this is something we can change, but we need something here and I think this is good enough at least for now until we figure out exactly how we should handle this?
Description
This is grab-back of various improvements for running Vercel tests:
run-local-test
so that it doesn't require you locally checkout repo to actually be inopennextjs-netlify
and instead just usepwd
to locate helper files we copy or useTests
N/A
Relevant links (GitHub issues, etc.) or a picture of cute animal