Closed thescientist13 closed 8 months ago
So was noticing this in the logs, maybe we just need to update our actions/setup-node version?
build (18) Node.js 12 actions are deprecated. Please update the following actions to use Node.js 16: actions/setup-node@v2. For more information see: https://github.blog/changelog/2022-09-22-github-actions-all-actions-will-begin-running-on-node16-instead-of-node12/.
Here is the linked blog article from GitHub re: the Node v12 deprecation https://github.blog/changelog/2022-09-22-github-actions-all-actions-will-begin-running-on-node16-instead-of-node12/
Eventually when the job is stopped or cancelled, it does seem to be related to a hung SSR test for static export, and only on Windows.
1) Build Greenwood With:
A Server Rendered Application (SSR) that is statically exported
"before all" hook in "A Server Rendered Application (SSR) that is statically exported":
Error: Timeout of 90000ms exceeded. For async tests and hooks, ensure "done()" is called; if returning a Promise, ensure it resolves. (D:\a\greenwood\greenwood\packages\cli\test\cases\build.default.ssr-static-export\build.default.ssr-static-export.spec.js)
at listOnTimeout (node:internal/timers:569:17)
at process.processTimers (node:internal/timers:512:7)
So maybe it is us?
Still not sure if environment or our code, seems like the release was working fine until things started breaking, even having to merge those last two just keep the flow of work moving π€ π€·ββοΈ https://github.com/ProjectEvergreen/greenwood/pull/1034/commits
So testing this particular spec out on my Windows VM and passed fine there once, but then running it a couple more times seemed to reproduce the issue of the process hanging? Will need to keep testing and playing around with things, but seems to be stalling consistently on the copy lifecycle. π€
Ok, so it actually seems it just doesn't exit the process in index.js? Added some logging and seems like all the lifecycles complete successfully.
Seems like we're seeing an issue now for just normal non Windows SSR specs!? π π
Not sure if fixed from something before or just after #1193 , but seems like we can close this one out once that one is merged! π
Type of Change
Other
Summary
Noticing that it seems our release related GitHub Actions have been failing inconsistently consistent (not any particular OS, not necessarily experimental or not).
Details
In particular the build just running on for hours, stalled out on the task, until it is eventually shut down.
So not even really a case of our tests / build failing, just seems more of an operation issue? (hopefully?)
As of the v0.28.0 release, these offending specs have been excluded for now, but maybe this and https://github.com/ProjectEvergreen/greenwood/issues/787 maybe we need to rethink our testing strategy? Maybe one without gallinago?
A number of failed attempts are linked to this issue so definitely want these to be ironed out for the 1.0 release so adding this to the 1.0 milestone so we don't forget.
Somewhat (un)related, this happens from time to time, so maybe just more of a case to rethink our testing? - https://github.com/ProjectEvergreen/greenwood/actions/runs/4648145282/jobs/8225644571