Closed roryabraham closed 1 month ago
Thinking about this some more, even if we had this in place it would not have caught the error we had. Because the URL we were using was wrong for the upload, there's no reason to think it would've been right on the verification step. So I'm going to just close this actually.
if we wanted to do this, I think we'd want it to be a manual check. Not sure how we'd fit that in as a deployer responsibility though
we decided to add this after all
The solution for this issue has been :rocket: deployed to production :rocket: in version 9.0.13-4 and is now subject to a 7-day regression period :calendar:. Here is the list of pull requests that resolve this issue:
If no regressions arise, payment will be issued on 2024-08-05. :confetti_ball:
Skipping the payment summary for this issue since all the assignees are employees or vendors. If this is incorrect, please manually add the payment summary SO.
Problem
Coming from here, we went a few days without noticing production deploys were broken.
Solution
We've addressed the root cause and I created https://github.com/rhysd/actionlint/issues/440 to try and prevent any similar type of error in the future.
However, let's also add an automated check that just confirms that the deploy completed as we expect, at least on web. In order to do this, we can just download the
version.json
file hosted on staging and/or production and verify that it matches what we expect after the deploy is done.