test-infra
test-infra copied to clipboard
kubernetes
Split prow config from prow codebase
We're getting close to being ready for the move. Items we know we need to accomplish is here:
- [x] reconfigure config updater for prow.k8s.io https://github.com/kubernetes/test-infra/pull/11781
- [x] move remaining config files to
config/and run theconfig-bootstrapperin case any changed had landed in between https://github.com/kubernetes/test-infra/pull/11784 - [x] ensure all references to original config locations are updated
- [x] move
prow/clustertoconfig/cluster - [x] ensure all auto-bump etc automation around
prow/clusteris updated - [ ] submit KEP to create repo
/cc @spiffxp @fejta @cjwagner @BenTheElder @krzyzacy
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle stale
Need to see what PR's modify config.yaml and plugin.yaml and decide if we want those to go through rebase fun
Rescoping based on what we thought was feasible for this quarter, moving to its own repo can be done as a followup
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle stale
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle stale
/remove-lifecycle stale
move prow/cluster to config/cluster ensure all auto-bump etc automation around prow/cluster is updated
sounds relatively easy, maybe we should do that. @cjwagner @fejta thoughts?
submit KEP to create repo
you actually just have to file an issue in k/org after getting consensus from the SIG on wanting it, no KEP necessary. the process is pretty lightweight. moving code is a little more involved to do cleanly.
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle stale
/remove-lifecycle stale
I think this is mostly done? pretty much everything is under config/.
submit KEP to create repo
we can actually do this with much less than a KEP, SIGs can sponsor a repo without going full KEP 🙃
I don't think there's any config lingering under prow/ now. some scripts / makefile stuff.
Yeah, the config is pretty much moved out of the prow directory. What would be needed to move prow into its own repo is:
- The repo itself
- Exporting the code to the new repo
- Setting up all prow-related jobs onto the new repo
- Announcing this in case other projects reference prow code (kubetest2?)
- Updating all references outside of /prow to prow
The point that is presumably most of the work is setting up the jobs on the new repo. It would be great to make this happen though, to better separate concerns and make certain tasks like updating dependencies easier :upside_down_face:
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle stale
Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.
If this issue is safe to close now please do so with /close.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta. /lifecycle rotten
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close.
Send feedback to sig-contributor-experience at kubernetes/community. /lifecycle stale
/remove-lifecycle stale /lifecycle frozen /priority important-longterm /area config /area jobs
Yeah, the config is pretty much moved out of the prow directory. What would be needed to move prow into its own repo is:
- The repo itself
We have https://github.com/kubernetes-sigs/prow which could host the codebase. Currently it just holds Prow docs but most of the code there sits under a /site folder so it should be straightforward to move code in there (I expect us to basically do mv prow/* <new-repo> and commit that).
- Setting up all prow-related jobs onto the new repo
I think at least for the jobs that publish new Prow images, they can be set up to use the code from the new repo and still publish into the same GCR locations, because the SHAs would be different so there wouldn't be any overlap. Once all of the presubmits and integration tests are set up to work on the new repo, I assume it would be OK to put a blockade on the old repo's /prow folder. Then the laborious task of setting up the Prow-related jobs can begin... and once that's done we can retire the old jobs.
I assume no one is currently working on this effort. But I am interested in this space.
This old issue got lost but https://github.com/kubernetes-sigs/prow exists now.