Need to check the difference between `SyncSet` and `Deployable`

[gyliu513]

@kuanf can you help check the difference between SyncSet and Deployable?

The SyncSet is a concept in Hive which is very similar with Deployable, here is the document https://github.com/openshift/hive/blob/master/docs/syncset.md

A sample SyncSet is as follows:

---
apiVersion: hive.openshift.io/v1alpha1
kind: SyncSet
metadata:
  name: mygroup
spec:
  clusterDeploymentRefs:
  - name: ClusterName

  resourceApplyMode: Upsert

  resources:
  - apiVersion: user.openshift.io/v1
    kind: Group
    metadata:
      name: mygroup
    users:
    - myuser

  patches:
  - kind: ConfigMap
    apiVersion: v1
    name: foo
    namespace: default
    applyMode: AlwaysApply
    patch: |-
      { "data": { "foo": "new-bar" } }
    patchType: merge

The SyncSet can not only distribute resources to different clusters, but also patch resources in different clusters.

FYI @mdelder @clyang82

[kuanf]

At very high level, they share very similar objective: facilitate resource management in managed clusters. I guess we all acknowledge the needs of having a resource sitting in one k8s without being executed by that k8s. In practice, they’re quite different.

• clustering model. sync set introduces clusterdeployment. clusterdeployment must sit with syncset in samenamespace. with that, I can image many clusterdeployment resource are created in user’s workspace and some of them may point to same cluster. In that case, it might not be easy to have a overview of all resources go to same cluster across namespaces. Deployable is leveraging clusterregistry and have only 1 cluster object for each cluster in cluster namespace. and user don’t put resource to cluster namespace themselves, but modify spec.placement to let controller create copy there.
• syncset/selectorsyncset accommodate more than 1 resource in 1 object, deployable only has 1.
• deployable has a spec.dependency to describe (createIfNotExist) relationship for this deployable to others, and other user can also leverage this field
• status. I didn’t find any thing in syncset status https://github.com/openshift/hive/blob/master/pkg/apis/hive/v1alpha1/syncset_types.go#L223. while deployable aggregates the status back to hub cluster. https://github.com/IBM/multicloud-operators-deployable/blob/master/pkg/apis/app/v1alpha1/deployable_types.go#L123
• deployable resource is very easy to interact with kubefed as you can find the placement and override section are the same. (placement is enhanced with a placementref in deployable).