Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
test/scripts: fail setup-osbuild-repo on bad repo
When testing with "pinned" osbuild commit IDs (defined in the Schutzfile), we set up a repository based on the commit ID that should exist for each commit in osbuild that had a (partially) successful CI run (the RPM jobs should at least run successfully). When a repository doesn't exist for whatever reason (the CI pipeline failed, the commit ID was copied incorrectly, etc), the setup script wouldn't fail and osbuild would be installed from the system repositories. This means that we're not testing what we think we are testing. In the best case our build jobs fail because the osbuild version in the repos isn't new enough to have all the features we're testing, but we might also end up having passing tests that aren't testing what we want, for example a change in osbuild internals that's on an open PR and we want to verify that none or our test image builds are broken. Make the setup-osbuild-repo script check that the repository metadata is reachable and exit with an error if it's not, making any test pipeline that requires it fail when necessary instead of silently skipping the repo configuration.
- Loading branch information