support to promote dependency resources automatically by using '-d=true'

[duanmengkk]

Signed-off-by: duanmeng [email protected]

What type of PR is this?
support to promote dependency resources automatically by using '-d=true'

What this PR does / why we need it:

If users use promote resource from member cluster to control plane,he may need promote relevant resources automatically.

Which issue(s) this PR fixes:
Fixes #1862

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

Support to use '--deps=true' or '-d true' to promote the relevant resources automatically`
example:
  kubectl karmada  promote deployments test -d=true  -c member1

[karmada-bot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
To complete the pull request process, please assign kevin-wangzefeng after the PR has been reviewed.
You can assign the PR to them by writing /assign @kevin-wangzefeng in a comment when ready.

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

Needs approval from an approver in each of these files:

• OWNERS

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

[duanmengkk]

/assign @lonelyCZ

[lonelyCZ]

Thanks @duanmengkk , I will review it ASAP. Could you give a sample of workflow including yaml and result? Perhaps, we can talk it in meeting tomorrow.

[duanmengkk]

1.Firstly,apply a deploy and config in membercluster. kubectl apply -f test-promote.yaml --context cluster30

---
kind: Deployment
apiVersion: apps/v1
metadata:
  name: test
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      k8s.kuboard.cn/name: test
  template:
    metadata:
      labels:
        k8s.kuboard.cn/name: test
    spec:
      volumes:
        - name: volume-2t8ic
          configMap:
            name: configdeps
      containers:
        - name: nginx
          image: 'nexus.cmss.com:8086/cnp/nginx:1.20.2'
          volumeMounts:
            - name: volume-2t8ic
              mountPath: /apps
          imagePullPolicy: IfNotPresent
      restartPolicy: Always
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: configdeps
  namespace: default
data:
  description: "test-config"

deploy and config is in member cluster now


2.Secondly,promote deploy to control plane with '-d=true', kubectl karmada promote deploy test -c cluster30 -d=true
the result is

and RB and works is applied fine.

3.Now deploy has been promoted into control plane,when delete deploy in control plane,kubectl delete -f test-promote.yaml,the deploy will be deleted in member cluster either.

[duanmengkk]

Any problems with this PR? @lonelyCZ

[lonelyCZ]

Thanks, I will review it ASAP.

[lonelyCZ]

Is it limited if we only use defaultinterpreter?

[lonelyCZ]

Is it limited if we only use defaultinterpreter?

[duanmengkk]

What do you think ? should we support customizedInterpreter in promote?

[lonelyCZ]

I think it is necessary because we can promote any resource, especially support CRD/CR.

Perhaps, we can listen some suggestions form @RainbowMango @mrlihanbo @GitHubxsy

[duanmengkk]

I think it is necessary because we can promote any resource, especially support CRD/CR.

Agree,But it looks a little difficult. As webhook handles resource in control plane,but promote handles resources in member cluster.
Let me doing some research on it.

[AvineshTripathi]

I think it is necessary because we can promote any resource, especially support CRD/CR.

Agree,But it looks a little difficult. As webhook handles resource in control plane,but promote handles resources in member cluster. Let me doing some research on it.

Hey @duanmengkk I was looking into the issue could you give a little bit more context on the above comment you made. Thanks already

[ikaven1024]

One question, when a CR is promoted, shall it's CRD to be promoted together?

[lonelyCZ]

One question, when a CR is promoted, shall it's CRD to be promoted together?

Perhaps, we should promote CRD firstly.

[lonelyCZ]

Hi @duanmengkk , Is there any progress on this PR?

[duanmengkk]

No,We can close this pr first. I have no idea about it yet,unless we introduce a webhook to member cluster.