-
-
Notifications
You must be signed in to change notification settings - Fork 275
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
cache breaks with private packages #222
Comments
This is quite weird. |
sure, will verify now. i also realized i left out some info, i'm updating the description appropriately (i opened this originally as a github support ticket because it involved secrets, actions, and dependabot). |
also yes, verified this works fine with the env variable:
|
Hi there. Setting up Dependabot with our CI in Github Actions we encountered the same issue. The build does not pass when Dependabot opens a PR but if I rerun the workflow it passes. One key difference I see is in the way the env variable |
The fact I think filing this to GitHub support nmd/or to https://github.com/actions/runner/issues would be the most best way to progress on this issue. |
I forgot to mention, in the OP's logs we see the same issues, failing logs has |
sigh we filed a ticket with github support, they said the problem must be with the 3rd party actions and sent us here. i'm going back to them about this, thanks for your help. |
Sorry about that, but they must be wrong, it is undiscutably an issue that secrets are not consistently set when running a job as shown by the logs and the resulting behavior. |
Given #230 I'm starting to wonder if |
unfortunately i tried that but it didn't work. also github confirmed this issue is expected because REASONS. apparently when dependabot runs there's no way to run the other actions as dependabot as well. |
whenever
Gemfile.lock
changes, the first action run always fails, but the second succeeds. we've seen this with prs fromdependabot
as well as prs opened by humans.if
Gemfile.lock
does not change then there's no issue.we have two different workflows with this issue - both are identical except for the last step (
bundle exec brakeman -q
vsbundle exec rubocop
)the first run will always fail:
but re-running manually always works:
failing logs: logs_588.zip
passing logs: logs_779.zip
i can't link to the runs because they're from a private repo, do you need anything else here?
The text was updated successfully, but these errors were encountered: