ROX-27716: Enable Konflux builds to replace GHA#15309
Conversation
|
Skipping CI for Draft Pull Request. |
This comment was marked as outdated.
This comment was marked as outdated.
|
Images are ready for the commit at 910484a. To use with deploy scripts, first |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #15309 +/- ##
==========================================
- Coverage 49.21% 49.21% -0.01%
==========================================
Files 2577 2577
Lines 189083 189083
==========================================
- Hits 93065 93062 -3
- Misses 88682 88684 +2
- Partials 7336 7337 +1
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
58d59cd to
e6a19aa
Compare
e6a19aa to
37d89bc
Compare
bb3c4e8 to
d53b86d
Compare
14c2a59 to
5c8c907
Compare
tommartensen
left a comment
There was a problem hiding this comment.
Few points for discussion:
- I find the tests easier to read if they are not table tests.
- Can we combine tests like "should tell Both when non-release tag pushed" for Konflux&GHA into one? It might make clearer what the env vars are set to in GHA&Konflux when a tag is pushed.
- Could the script, test and hold file live in a separate directory in
scripts/ci?
Deal.
You suggest to change the grouping per use-case (so that groups are "release tag", "non-release tag", "release branch push", "non-release branch push", "PR to release branch", etc.) instead of grouping per CI ("Konflux" and "GHA"). Right? I can easily make it. Just need to make sure I understand correctly.
Since the thing is called from both GHA and Konflux I thought pacing it in scripts/ci is ok. I can't come up with a better location or a subdirectory name. Please suggest. |
Yes, that was my intention.
I don't have a good suggestion - let's put leave it there then! |
|
Reordered tests, PTAL. |
build in favor of Konflux build.
to get Konflux image tag suffix suppression
When there's only one matrix item (RHACS_BRANDING) and when it gets excluded (with `exclude`), the corresponding job still runs but in no-matrix mode. It's weird but that's the way it works. `build-and-push-operator` seems to be the only job like this and for it the default branding STACKROX_BRANDING gets assumed which results in building images tagged for quay.io/stackrox-io. Subsequently, any job steps that try to work with quay.io/rhacs-eng images fail and so I disable them with an additional clause in the `if` condition.
Only moved test functions and updated some comments. No change to tests themselves.
c53914b to
910484a
Compare
I reverted changes from #15309 except for changes in `.tekton/` directory (where the new task was used because that'll be updated in this same PR). My sequence was: ``` git revert 0309e52 git reset --soft HEAD~ git restore --staged .tekton git commit ``` No other changes. This makes sure `should-konflux-replace-gha-build.*` stuff disappears as well as its use in GHA.
I reverted changes from #15309 except for changes in `.tekton/` directory (where the new task was used because that'll be updated in this same PR). My sequence was: ``` git revert 0309e52 git reset --soft HEAD~ git restore --staged .tekton git commit ``` No other changes. This makes sure `should-konflux-replace-gha-build.*` stuff disappears as well as its use in GHA.
I reverted changes from #15309 except for changes in `.tekton/` directory (where the new task was used because that'll be updated in this same PR). My sequence was: ``` git revert 0309e52 git reset --soft HEAD~ git restore --staged .tekton git commit ``` No other changes. This makes sure `should-konflux-replace-gha-build.*` stuff disappears as well as its use in GHA.
Description
For release builds, for
quay.io/rhacs-engimages.Paired with stackrox/konflux-tasks#55
User-facing documentation
Testing and quality
Automated testing
No change.
How I validated my change
See stackrox/konflux-tasks#55