docs(proposal): Migration Rollback Protection

[CharlesQQ]

What type of PR is this?

What this PR does / why we need it:

Which issue(s) this PR fixes:
part of #4709

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

NONE

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 34.85%. Comparing base (721107d) to head (fb99665).
Report is 8 commits behind head on master.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5101      +/-   ##
==========================================
- Coverage   34.87%   34.85%   -0.03%     
==========================================
  Files         645      645              
  Lines       44833    44861      +28     
==========================================
- Hits        15637    15635       -2     
- Misses      27991    28021      +30     
  Partials     1205     1205              

Flag	Coverage Δ
unittests	34.85% <ø> (-0.03%)	⬇️

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

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

[XiShanYongYe-Chang]

Thanks a lot @CharlesQQ

Hi, guys @RainbowMango @chaunceyjiang @whitewindmills @zhy76 @buji-code @lcw2
We can discuss here to improve the resource deletion policy.

[whitewindmills]

/assign

[whitewindmills]

does this annotation have a knock-on effect on dependent resources?―that is, one adds such an annotation to his Deployment, will the annotation affect its dependent resources(like ConfigMap)?

[CharlesQQ]

I see, what do the author @CharlesQQ thinik of #5101 (comment)?

Extended by Annotation or Extend the fields of Work Choose one, I'm more tend to solution 1.

[chaosi-zju]

I see, what do the author @CharlesQQ thinik of #5101 (comment)?

Extended by Annotation or Extend the fields of Work Choose one, I'm more tend to solution 1.

got it, solution 1（+1）

[RainbowMango]

For now, during the process of deleting the workload, the execution-controller unmarshals the manifest already, does this approach adds extra overhead?

karmada/pkg/controllers/execution/execution_controller.go

Lines 165 to 182 in b27e669

	func (c *Controller) tryDeleteWorkload(ctx context.Context, clusterName string, work *workv1alpha1.Work) error {
	for _, manifest := range work.Spec.Workload.Manifests {
	workload := &unstructured.Unstructured{}
	err := workload.UnmarshalJSON(manifest.Raw)
	if err != nil {
	klog.Errorf("Failed to unmarshal workload, error is: %v", err)
	return err
	}
	err = c.ObjectWatcher.Delete(ctx, clusterName, workload)
	if err != nil {
	klog.Errorf("Failed to delete resource in the given member cluster %v, err is %v", clusterName, err)
	return err
	}
	}
	return nil
	}

[XiShanYongYe-Chang]

Good point. For this approach, no extra parsing overhead is required.

[RainbowMango]

Suggested change

	---
	title: Use Cascading Deletion in Karmada
	authors:
	- "@CharlesQQ"
	reviewers:
	- "@robot"
	- TBD
	approvers:
	- "@robot"
	- TBD
	creation-date: 2024-07-01
	---
	title: Use Cascading Deletion in Karmada
	authors:
	- "@CharlesQQ"
	reviewers:
	- "@robot"
	- TBD
	approvers:
	- "@robot"
	- TBD
	creation-date: 2024-07-01
	---

Please update the reviewers by adding the guys who reviewed this proposal.

[RainbowMango]

How about naming this feature as Migration Rollback Protection?

Forget cascading to avoid misleading.

[XiShanYongYe-Chang]

+1

[XiShanYongYe-Chang]

cc @RainbowMango
Help take a review again.

[RainbowMango]

Please remove all Chinese characters.

[RainbowMango]

Suggested change

	The current behavior of the Karmada system is that when a user deletes resources from the Karmada control plane, the distributed resources in the member clusters are also deleted synchronously. However, in certain scenarios, such as workload migration rollout scenarios, users may wish to retain the workloads in the member clusters.
	The current behavior of the Karmada system is that when a user deletes resources from the Karmada control plane, the distributed resources in the member clusters are also deleted synchronously. However, in certain scenarios, such as workload migration scenarios, users may wish to retain the workloads in the member clusters.

[RainbowMango]

/lgtm
/approve

[karmada-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: RainbowMango

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

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [RainbowMango]

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