Rename TektonScheduler to TektonKueue

[waveywaves]

Changes

Rename the Operator component API from TektonScheduler to TektonKueue,
including TektonConfig.spec.kueue, controller wiring, generated clients and
CRDs, RBAC, Helm and OperatorHub manifests, tests, and documentation.

Retain the deprecated TektonScheduler CRD and TektonConfig.spec.scheduler
field during a compatibility window. Pre-upgrade reconciliation migrates
existing configuration and legacy resources to TektonKueue; explicitly
configured spec.kueue values take precedence. It also removes deprecated
InstallerSets and RBAC, waits for legacy resource deletion to complete, and
uses delayed requeues for expected migration steps instead of controller error
backoff.

The branch is rebased onto current main and generated CRD/Helm artifacts are
refreshed against Tekton Kueue v0.4.0. Repository and release-fetch references
use tektoncd/tekton-kueue. Go imports retain
github.com/konflux-ci/tekton-kueue because v0.4.0 still declares that module
path.

Automated migration coverage includes disabled Scheduler configuration,
enabled configuration without a queue override, a changed queue name,
explicit spec.kueue precedence, InstallerSet/RBAC cleanup, and finalizer
handling.

A local Kind upgrade matrix from Operator v0.81.1 on Kubernetes v1.36.1
covered Scheduler disabled, Scheduler enabled with defaults, and Scheduler
enabled with queueName: migration-test-queue. It verified that:

• disabled configuration migrated without creating a TektonKueue operand;
• enabled default configuration retained the default pipelines-queue;
• the custom migration-test-queue value was preserved;
• enabled TektonKueue resources reached Ready;
• legacy Scheduler InstallerSets, RBAC, and resources were removed;
• both old and new CRDs remained served during the compatibility window; and
• TektonConfig returned to Ready.

A raw-manifest upgrade on OpenShift 4.20.33 from Operator v0.81.1 with
queueName: migration-test-queue independently verified the same migration,
cleanup, queue preservation, and Ready states. All cluster operators remained
available and non-degraded.

Submitter Checklist

These are the criteria that every PR should meet, please check them off as you
review them:

• Run make test lint before submitting a PR
• Includes tests (if functionality changed/added)
• Includes docs (if user facing)
• Commit messages follow commit message best practices

See the contribution guide for more details.

Release Notes

`TektonScheduler` is deprecated in favor of `TektonKueue`, and
`TektonConfig.spec.scheduler` is deprecated in favor of
`TektonConfig.spec.kueue`. Existing configuration is migrated automatically
during pre-upgrade reconciliation.

Action required: rename scheduler image overrides to `IMAGE_KUEUE_MANAGER` and
`IMAGE_KUEUE_WEBHOOK`.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 27.26%. Comparing base (f09782c) to head (bbf14cf).
⚠️ Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3809      +/-   ##
==========================================
+ Coverage   26.95%   27.26%   +0.30%     
==========================================
  Files         467      475       +8     
  Lines       25170    25428     +258     
==========================================
+ Hits         6785     6932     +147     
- Misses      17648    17745      +97     
- Partials      737      751      +14     

Flag	Coverage Δ
unit-tests	27.26% <ø> (+0.30%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[pramodbindal]

As you mentioned this is breaking change. we need to provide some backward compatibility.

Something like. If you can keep the dummy TektonScheduler for some releases so existing TektonScheduler field does not create any issue and if present we need to move the config to TektonKueue.
This can be done in pre-upgrade Remove the Scheduler so after one release all customers will be moved to TektonScheduler from TektonKueue

[anithapriyanatarajan]

@waveywaves - do we wait until the tekton-kueue project is moved to tektoncd org?

[vdemeester]

@waveywaves - do we wait until the tekton-kueue project is moved to tektoncd org?

for what is worth, it has moved in tektoncd org today 😛

[anithapriyanatarajan]

@waveywaves - do we wait until the tekton-kueue project is moved to tektoncd org?

for what is worth, it has moved in tektoncd org today 😛

Thank you 😄 @waveywaves could you update the import pkg references and components.yaml references. Thank you

[khrm]

Did you run this in OpenShift? Is it working?

[pramodbindal]

@waveywaves
Can you upload the upgrade testing results with following settings in old cluster

1. When Scheduler is not enabled
2. When Scheduler is enabled
3. When Scheduler is enabled and some other param (like default queue name ) is also changed

[waveywaves]

@pramodbindal @khrm I rebased the PR onto current main and expanded the upgrade handling and validation.

The compatibility path now retains the deprecated CRD and spec.scheduler field, migrates existing configuration to TektonKueue, preserves an explicit queue name, removes legacy InstallerSets and RBAC, waits for legacy resource deletion, and keeps spec.kueue authoritative when both fields are present.

Upgrade matrix from Operator v0.81.1 to this PR head (5a5070cc) on Kind/Kubernetes v1.36.1:

Old configuration	Result after upgrade
Scheduler disabled	spec.kueue.disabled=true; no TektonKueue operand created; legacy RBAC absent; both CRDs served; TektonConfig Ready
Scheduler enabled with defaults	spec.kueue.disabled=false; default pipelines-queue preserved; TektonKueue Ready; legacy CR/InstallerSets/RBAC removed; both CRDs served; TektonConfig Ready
Scheduler enabled with queueName: migration-test-queue	spec.kueue.disabled=false; queue preserved; TektonKueue Ready; legacy CR/InstallerSets/RBAC removed; both CRDs served; TektonConfig Ready; full local apply-to-ready upgrade completed in 86 seconds

The upgrade runs also exposed two issues that are now fixed and covered: a missing delete permission for cert-manager resources during legacy InstallerSet cleanup, and expected migration steps returning ordinary errors, which allowed a long controller retry backoff. Migration now uses fixed delayed requeues.

Local make test, make lint, and generated CRD/Helm synchronization pass. GitHub build, test, lint, generated-code, multi-arch, security analysis, and Kubernetes e2e checks pass.

I also ran the representative custom-queue upgrade through the raw OpenShift manifests on OpenShift 4.20.33/Kubernetes v1.33.13. The apply-to-ready phase completed in 132 seconds: the queue was preserved, TektonKueue and TektonConfig became Ready, legacy InstallerSets/RBAC/resource were removed, both CRDs remained served, all operand pods were healthy, and no cluster operator was degraded or unavailable.

[waveywaves]

cc @jkhelil

[waveywaves]

@khrm @pramodbindal Could you please take another look?

[khrm]

I think we shoudl merge this since it was tested in OpenShift.

[khrm]

/approve

@waveywaves Please rebase.

[tekton-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: khrm

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [khrm]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[waveywaves]

@jkhelil can you check again ? I have rebased and updated the PR PTAL 🙇

[jkhelil]

/lgtm