PersistentVolumeClaim MetaData can not be decoded correctly

[hzyfox]

Environment

Kubectl Version

Client Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.4", GitCommit:"c96aede7b5205121079932896c4ad89bb93260af", GitTreeState:"clean", BuildDate:"2020-06-18T02:59:13Z", GoVersion:"go1.14.3", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"20+", GitVersion:"v1.20.4-80+89e0897d2cb807", GitCommit:"89e0897d2cb8073fbb8f700258573f1478d4826a", GitTreeState:"clean", BuildDate:"2021-11-22T03:53:35Z", GoVersion:"go1.15.8", Compiler:"gc", Platform:"linux/amd64"}

Kubernetes Version (Kind Cluster)

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
  - role: control-plane
    image: kindest/node:v1.20.7@sha256:cbeaf907fc78ac97ce7b625e4bf0de16e3ea725daf6b04f930bd14c67c671ff9
  - role: worker
    image: kindest/node:v1.20.7@sha256:cbeaf907fc78ac97ce7b625e4bf0de16e3ea725daf6b04f930bd14c67c671ff9
  - role: worker
    image: kindest/node:v1.20.7@sha256:cbeaf907fc78ac97ce7b625e4bf0de16e3ea725daf6b04f930bd14c67c671ff9

Kubebuilder Version

Version: main.version{KubeBuilderVersion:"3.1.0", KubernetesVendor:"1.19.2", GitCommit:"92e0349ca7334a0a8e5e499da4fb077eb524e94a", BuildDate:"2021-05-27T17:54:28Z", GoOs:"darwin", GoArch:"amd64"}

Os

Macos Big Sur 11.6

I use kubebuilder to define my own CRD like below, and it contains VolumeClaimTemplates filed which the type is []coreV1.PersistentVolumeClaim

package v1alpha1
import (
	apps "k8s.io/api/apps/v1"
	coreV1 "k8s.io/api/core/v1"
	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
	"k8s.io/apimachinery/pkg/util/intstr"
)

type DatabaseSetSpec struct {
        ...
        // +optional
	VolumeClaimTemplates []coreV1.PersistentVolumeClaim `json:"volumeClaimTemplates,omitempty" protobuf:"bytes,4,rep,name=volumeClaimTemplates"`
}

// +kubebuilder:object:root=true
// +kubebuilder:subresource:status
// +kubebuilder:resource:shortName=ami-dbs

// DatabaseSet is the Schema for the databasesets API
type DatabaseSet struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata,omitempty"`

	Spec   DatabaseSetSpec   `json:"spec,omitempty"`
	Status DatabaseSetStatus `json:"status,omitempty"`
}

// DatabaseSetStatus defines the observed state of DatabaseSet
type DatabaseSetStatus struct {
    ...
}

//+kubebuilder:object:root=true

// DatabaseSetList contains a list of DatabaseSet
type DatabaseSetList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []DatabaseSet `json:"items"`
}

func init() {
	SchemeBuilder.Register(&DatabaseSet{}, &DatabaseSetList{})
}

But when I apply the CR like the below, I found that the metadata filed is empty

apiVersion: apps.analyticdb.aliyun.com/v1alpha1
kind: DatabaseSet
metadata:
  name: databaseset-sample
spec: 
  ...
  volumeClaimTemplates:
    - metadata:
        name: pvc-test
      spec:
        accessModes: [ "ReadWriteOnce" ]
        storageClassName: "manual"
        resources:
          requests:
            storage: 3Gi

Here is the yaml which get from the k8s ectd, it cound be found that the metadata of the volumeClaimTemplates is empty.

apiVersion: apps.analyticdb.aliyun.com/v1alpha1
kind: DatabaseSet
metadata:
  creationTimestamp: "2021-12-24T09:46:22Z"
  generation: 1
  name: databaseset-sample
  namespace: default
  resourceVersion: "98727469"
  uid: e64107f2-7a4b-473b-9275-39ab5e2e88dc
spec:
  ...
  volumeClaimTemplates:
  - metadata: {}
    spec:
      accessModes:
      - ReadWriteOnce
      resources:
        requests:
          storage: 3Gi
      storageClassName: manual

Does anyone know why？

And when I mark the volumeclaimtemplate field with the below comment

// +kubebuilder:pruning:PreserveUnknownFields
// +kubebuilder:validation:Schemaless

metada can be decoed correctly

[camilamacedo86]

Using the following markers will prevent pruning and enable schema-less validation, which might solve the issue:

```go
// +kubebuilder:pruning:PreserveUnknownFields
// +kubebuilder:validation:Schemaless
```

By applying these markers, you inform Kubernetes to retain all fields (even those it doesn't recognize) and not to enforce a schema on this particular field.

Ensure to regenerate the CRD and apply it after adding these markers. For further info see: https://book.kubebuilder.io/reference/markers

I am closing this one however, if you need to help about this one feel free to reopen