Merge pull request #9015 from Lyndon-Li/vgdp-ms-pvb-controller

VGDP MS PVB controller

changelogs/unreleased/9015-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #8958, add VGDP MS PVB controller

config/crd/v1/bases/velero.io_podvolumebackups.yaml

@@ -15,38 +15,41 @@ spec:
   scope: Namespaced
   versions:
   - additionalPrinterColumns:
-    - description: Pod Volume Backup status such as New/InProgress
+    - description: PodVolumeBackup status such as New/InProgress
       jsonPath: .status.phase
       name: Status
       type: string
-    - description: Time when this backup was started
+    - description: Time duration since this PodVolumeBackup was started
       jsonPath: .status.startTimestamp
-      name: Created
+      name: Started
       type: date
-    - description: Namespace of the pod containing the volume to be backed up
+    - description: Completed bytes
-      jsonPath: .spec.pod.namespace
+      format: int64
-      name: Namespace
+      jsonPath: .status.progress.bytesDone
-      type: string
+      name: Bytes Done
-    - description: Name of the pod containing the volume to be backed up
+      type: integer
-      jsonPath: .spec.pod.name
+    - description: Total bytes
-      name: Pod
+      format: int64
-      type: string
+      jsonPath: .status.progress.totalBytes
-    - description: Name of the volume to be backed up
+      name: Total Bytes
-      jsonPath: .spec.volume
+      type: integer
-      name: Volume
-      type: string
-    - description: The type of the uploader to handle data transfer
-      jsonPath: .spec.uploaderType
-      name: Uploader Type
-      type: string
     - description: Name of the Backup Storage Location where this backup should be
         stored
       jsonPath: .spec.backupStorageLocation
       name: Storage Location
       type: string
-    - jsonPath: .metadata.creationTimestamp
+    - description: Time duration since this PodVolumeBackup was created
+      jsonPath: .metadata.creationTimestamp
       name: Age
       type: date
+    - description: Name of the node where the PodVolumeBackup is processed
+      jsonPath: .status.node
+      name: Node
+      type: string
+    - description: The type of the uploader to handle data transfer
+      jsonPath: .spec.uploaderType
+      name: Uploader
+      type: string
     name: v1
     schema:
       openAPIV3Schema:
@@ -170,6 +173,13 @@ spec:
           status:
             description: PodVolumeBackupStatus is the current status of a PodVolumeBackup.
             properties:
+              acceptedTimestamp:
+                description: |-
+                  AcceptedTimestamp records the time the pod volume backup is to be prepared.
+                  The server's time is used for AcceptedTimestamp
+                format: date-time
+                nullable: true
+                type: string
               completionTimestamp:
                 description: |-
                   CompletionTimestamp records the time a backup was completed.
@@ -190,7 +200,11 @@ spec:
                 description: Phase is the current state of the PodVolumeBackup.
                 enum:
                 - New
+                - Accepted
+                - Prepared
                 - InProgress
+                - Canceling
+                - Canceled
                 - Completed
                 - Failed
                 type: string

pkg/apis/velero/v1/labels_annotations.go

@@ -105,6 +105,9 @@ const (
 	// defaultVGSLabelKey is the default label key used to group PVCs under a VolumeGroupSnapshot
 	DefaultVGSLabelKey = "velero.io/volume-group"
 
+	// PVBLabel is the label key used to identify the pvb for pvb pod
+	PVBLabel = "velero.io/pod-volume-backup"
+
 	// PVRLabel is the label key used to identify the pvb for pvr pod
 	PVRLabel = "velero.io/pod-volume-restore"
 )

pkg/apis/velero/v1/pod_volume_backup_types.go

@@ -64,12 +64,16 @@ type PodVolumeBackupSpec struct {
 }
 
 // PodVolumeBackupPhase represents the lifecycle phase of a PodVolumeBackup.
-// +kubebuilder:validation:Enum=New;InProgress;Completed;Failed
+// +kubebuilder:validation:Enum=New;Accepted;Prepared;InProgress;Canceling;Canceled;Completed;Failed
 type PodVolumeBackupPhase string
 
 const (
 	PodVolumeBackupPhaseNew        PodVolumeBackupPhase = "New"
+	PodVolumeBackupPhaseAccepted   PodVolumeBackupPhase = "Accepted"
+	PodVolumeBackupPhasePrepared   PodVolumeBackupPhase = "Prepared"
 	PodVolumeBackupPhaseInProgress PodVolumeBackupPhase = "InProgress"
+	PodVolumeBackupPhaseCanceling  PodVolumeBackupPhase = "Canceling"
+	PodVolumeBackupPhaseCanceled   PodVolumeBackupPhase = "Canceled"
 	PodVolumeBackupPhaseCompleted  PodVolumeBackupPhase = "Completed"
 	PodVolumeBackupPhaseFailed     PodVolumeBackupPhase = "Failed"
 )
@@ -113,20 +117,27 @@ type PodVolumeBackupStatus struct {
 	// about the backup operation.
 	// +optional
 	Progress shared.DataMoveOperationProgress `json:"progress,omitempty"`
+
+	// AcceptedTimestamp records the time the pod volume backup is to be prepared.
+	// The server's time is used for AcceptedTimestamp
+	// +optional
+	// +nullable
+	AcceptedTimestamp *metav1.Time `json:"acceptedTimestamp,omitempty"`
 }
 
 // TODO(2.0) After converting all resources to use the runttime-controller client,
 // the genclient and k8s:deepcopy markers will no longer be needed and should be removed.
 // +genclient
 // +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
-// +kubebuilder:printcolumn:name="Status",type="string",JSONPath=".status.phase",description="Pod Volume Backup status such as New/InProgress"
+// +kubebuilder:storageversion
-// +kubebuilder:printcolumn:name="Created",type="date",JSONPath=".status.startTimestamp",description="Time when this backup was started"
+// +kubebuilder:printcolumn:name="Status",type="string",JSONPath=".status.phase",description="PodVolumeBackup status such as New/InProgress"
-// +kubebuilder:printcolumn:name="Namespace",type="string",JSONPath=".spec.pod.namespace",description="Namespace of the pod containing the volume to be backed up"
+// +kubebuilder:printcolumn:name="Started",type="date",JSONPath=".status.startTimestamp",description="Time duration since this PodVolumeBackup was started"
-// +kubebuilder:printcolumn:name="Pod",type="string",JSONPath=".spec.pod.name",description="Name of the pod containing the volume to be backed up"
+// +kubebuilder:printcolumn:name="Bytes Done",type="integer",format="int64",JSONPath=".status.progress.bytesDone",description="Completed bytes"
-// +kubebuilder:printcolumn:name="Volume",type="string",JSONPath=".spec.volume",description="Name of the volume to be backed up"
+// +kubebuilder:printcolumn:name="Total Bytes",type="integer",format="int64",JSONPath=".status.progress.totalBytes",description="Total bytes"
-// +kubebuilder:printcolumn:name="Uploader Type",type="string",JSONPath=".spec.uploaderType",description="The type of the uploader to handle data transfer"
 // +kubebuilder:printcolumn:name="Storage Location",type="string",JSONPath=".spec.backupStorageLocation",description="Name of the Backup Storage Location where this backup should be stored"
-// +kubebuilder:printcolumn:name="Age",type="date",JSONPath=".metadata.creationTimestamp"
+// +kubebuilder:printcolumn:name="Age",type="date",JSONPath=".metadata.creationTimestamp",description="Time duration since this PodVolumeBackup was created"
+// +kubebuilder:printcolumn:name="Node",type="string",JSONPath=".status.node",description="Name of the node where the PodVolumeBackup is processed"
+// +kubebuilder:printcolumn:name="Uploader",type="string",JSONPath=".spec.uploaderType",description="The type of the uploader to handle data transfer"
 // +kubebuilder:object:root=true
 // +kubebuilder:object:generate=true
 

pkg/apis/velero/v1/zz_generated.deepcopy.go

@@ -1043,6 +1043,10 @@ func (in *PodVolumeBackupStatus) DeepCopyInto(out *PodVolumeBackupStatus) {
 		*out = (*in).DeepCopy()
 	}
 	out.Progress = in.Progress
+	if in.AcceptedTimestamp != nil {
+		in, out := &in.AcceptedTimestamp, &out.AcceptedTimestamp
+		*out = (*in).DeepCopy()
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PodVolumeBackupStatus.

pkg/builder/pod_volume_backup_builder.go

@@ -119,3 +119,33 @@ func (b *PodVolumeBackupBuilder) Annotations(annotations map[string]string) *Pod
 	b.object.Annotations = annotations
 	return b
 }
+
+// Cancel sets the PodVolumeBackup's Cancel.
+func (b *PodVolumeBackupBuilder) Cancel(cancel bool) *PodVolumeBackupBuilder {
+	b.object.Spec.Cancel = cancel
+	return b
+}
+
+// AcceptedTimestamp sets the PodVolumeBackup's AcceptedTimestamp.
+func (b *PodVolumeBackupBuilder) AcceptedTimestamp(acceptedTimestamp *metav1.Time) *PodVolumeBackupBuilder {
+	b.object.Status.AcceptedTimestamp = acceptedTimestamp
+	return b
+}
+
+// Finalizers sets the PodVolumeBackup's Finalizers.
+func (b *PodVolumeBackupBuilder) Finalizers(finalizers []string) *PodVolumeBackupBuilder {
+	b.object.Finalizers = finalizers
+	return b
+}
+
+// Message sets the PodVolumeBackup's Message.
+func (b *PodVolumeBackupBuilder) Message(msg string) *PodVolumeBackupBuilder {
+	b.object.Status.Message = msg
+	return b
+}
+
+// OwnerReference sets the PodVolumeBackup's OwnerReference.
+func (b *PodVolumeBackupBuilder) OwnerReference(ref metav1.OwnerReference) *PodVolumeBackupBuilder {
+	b.object.OwnerReferences = append(b.object.OwnerReferences, ref)
+	return b
+}

pkg/cmd/cli/nodeagent/server.go

@@ -48,7 +48,6 @@ import (
 
 	snapshotv1client "github.com/kubernetes-csi/external-snapshotter/client/v7/clientset/versioned"
 
-	"github.com/vmware-tanzu/velero/internal/credentials"
 	velerov1api "github.com/vmware-tanzu/velero/pkg/apis/velero/v1"
 	velerov2alpha1api "github.com/vmware-tanzu/velero/pkg/apis/velero/v2alpha1"
 	"github.com/vmware-tanzu/velero/pkg/buildinfo"
@@ -60,7 +59,6 @@ import (
 	"github.com/vmware-tanzu/velero/pkg/datapath"
 	"github.com/vmware-tanzu/velero/pkg/metrics"
 	"github.com/vmware-tanzu/velero/pkg/nodeagent"
-	"github.com/vmware-tanzu/velero/pkg/repository"
 	"github.com/vmware-tanzu/velero/pkg/util/filesystem"
 	"github.com/vmware-tanzu/velero/pkg/util/kube"
 	"github.com/vmware-tanzu/velero/pkg/util/logging"
@@ -282,30 +280,6 @@ func (s *nodeAgentServer) run() {
 
 	s.logger.Info("Starting controllers")
 
-	credentialFileStore, err := credentials.NewNamespacedFileStore(
-		s.mgr.GetClient(),
-		s.namespace,
-		credentials.DefaultStoreDirectory(),
-		filesystem.NewFileSystem(),
-	)
-	if err != nil {
-		s.logger.Fatalf("Failed to create credentials file store: %v", err)
-	}
-
-	credSecretStore, err := credentials.NewNamespacedSecretStore(s.mgr.GetClient(), s.namespace)
-	if err != nil {
-		s.logger.Fatalf("Failed to create secret file store: %v", err)
-	}
-
-	credentialGetter := &credentials.CredentialGetter{FromFile: credentialFileStore, FromSecret: credSecretStore}
-	repoEnsurer := repository.NewEnsurer(s.mgr.GetClient(), s.logger, s.config.resourceTimeout)
-	pvbReconciler := controller.NewPodVolumeBackupReconciler(s.mgr.GetClient(), s.kubeClient, s.dataPathMgr, repoEnsurer,
-		credentialGetter, s.nodeName, s.mgr.GetScheme(), s.metrics, s.logger)
-
-	if err := pvbReconciler.SetupWithManager(s.mgr); err != nil {
-		s.logger.Fatal(err, "unable to create controller", "controller", constant.ControllerPodVolumeBackup)
-	}
-
 	var loadAffinity *kube.LoadAffinity
 	if s.dataPathConfigs != nil && len(s.dataPathConfigs.LoadAffinity) > 0 {
 		loadAffinity = s.dataPathConfigs.LoadAffinity[0]
@@ -328,7 +302,12 @@ func (s *nodeAgentServer) run() {
 		}
 	}
 
-	if err = controller.NewPodVolumeRestoreReconciler(s.mgr.GetClient(), s.mgr, s.kubeClient, s.dataPathMgr, s.nodeName, s.config.dataMoverPrepareTimeout, s.config.resourceTimeout, podResources, s.logger).SetupWithManager(s.mgr); err != nil {
+	pvbReconciler := controller.NewPodVolumeBackupReconciler(s.mgr.GetClient(), s.mgr, s.kubeClient, s.dataPathMgr, s.nodeName, s.config.dataMoverPrepareTimeout, s.config.resourceTimeout, podResources, s.metrics, s.logger)
+	if err := pvbReconciler.SetupWithManager(s.mgr); err != nil {
+		s.logger.Fatal(err, "unable to create controller", "controller", constant.ControllerPodVolumeBackup)
+	}
+
+	if err := controller.NewPodVolumeRestoreReconciler(s.mgr.GetClient(), s.mgr, s.kubeClient, s.dataPathMgr, s.nodeName, s.config.dataMoverPrepareTimeout, s.config.resourceTimeout, podResources, s.logger).SetupWithManager(s.mgr); err != nil {
 		s.logger.WithError(err).Fatal("Unable to create the pod volume restore controller")
 	}
 
@@ -347,7 +326,7 @@ func (s *nodeAgentServer) run() {
 		s.logger,
 		s.metrics,
 	)
-	if err = dataUploadReconciler.SetupWithManager(s.mgr); err != nil {
+	if err := dataUploadReconciler.SetupWithManager(s.mgr); err != nil {
 		s.logger.WithError(err).Fatal("Unable to create the data upload controller")
 	}
 
@@ -358,7 +337,7 @@ func (s *nodeAgentServer) run() {
 	}
 
 	dataDownloadReconciler := controller.NewDataDownloadReconciler(s.mgr.GetClient(), s.mgr, s.kubeClient, s.dataPathMgr, restorePVCConfig, podResources, s.nodeName, s.config.dataMoverPrepareTimeout, s.logger, s.metrics)
-	if err = dataDownloadReconciler.SetupWithManager(s.mgr); err != nil {
+	if err := dataDownloadReconciler.SetupWithManager(s.mgr); err != nil {
 		s.logger.WithError(err).Fatal("Unable to create the data download controller")
 	}
 

pkg/controller/data_upload_controller_test.go

@@ -312,30 +312,29 @@ func (f *fakeSnapshotExposer) DiagnoseExpose(context.Context, corev1api.ObjectRe
 func (f *fakeSnapshotExposer) CleanUp(context.Context, corev1api.ObjectReference, string, string) {
 }
 
-type fakeDataUploadFSBR struct {
+type fakeFSBR struct {
-	du         *velerov2alpha1api.DataUpload
 	kubeClient kbclient.Client
 	clock      clock.WithTickerAndDelayedExecution
 	initErr    error
 	startErr   error
 }
 
-func (f *fakeDataUploadFSBR) Init(ctx context.Context, param any) error {
+func (f *fakeFSBR) Init(ctx context.Context, param any) error {
 	return f.initErr
 }
 
-func (f *fakeDataUploadFSBR) StartBackup(source datapath.AccessPoint, uploaderConfigs map[string]string, param any) error {
+func (f *fakeFSBR) StartBackup(source datapath.AccessPoint, uploaderConfigs map[string]string, param any) error {
 	return f.startErr
 }
 
-func (f *fakeDataUploadFSBR) StartRestore(snapshotID string, target datapath.AccessPoint, uploaderConfigs map[string]string) error {
+func (f *fakeFSBR) StartRestore(snapshotID string, target datapath.AccessPoint, uploaderConfigs map[string]string) error {
 	return nil
 }
 
-func (b *fakeDataUploadFSBR) Cancel() {
+func (b *fakeFSBR) Cancel() {
 }
 
-func (b *fakeDataUploadFSBR) Close(ctx context.Context) {
+func (b *fakeFSBR) Close(ctx context.Context) {
 }
 
 func TestReconcile(t *testing.T) {
@@ -651,8 +650,7 @@ func TestReconcile(t *testing.T) {
 			}
 
 			datapath.MicroServiceBRWatcherCreator = func(kbclient.Client, kubernetes.Interface, manager.Manager, string, string, string, string, string, string, datapath.Callbacks, logrus.FieldLogger) datapath.AsyncBR {
-				return &fakeDataUploadFSBR{
+				return &fakeFSBR{
-					du:         test.du,
 					kubeClient: r.client,
 					clock:      r.Clock,
 					initErr:    test.fsBRInitErr,

pkg/podvolume/backupper.go

@@ -169,7 +169,8 @@ func newBackupper(
 				}
 
 				if pvb.Status.Phase != velerov1api.PodVolumeBackupPhaseCompleted &&
-					pvb.Status.Phase != velerov1api.PodVolumeBackupPhaseFailed {
+					pvb.Status.Phase != velerov1api.PodVolumeBackupPhaseFailed &&
+					pvb.Status.Phase != velerov1api.PodVolumeBackupPhaseCanceled {
 					return
 				}
 
@@ -179,7 +180,8 @@ func newBackupper(
 					existPVB, ok := existObj.(*velerov1api.PodVolumeBackup)
 					// the PVB in the indexer is already in final status, no need to call WaitGroup.Done()
 					if ok && (existPVB.Status.Phase == velerov1api.PodVolumeBackupPhaseCompleted ||
-						existPVB.Status.Phase == velerov1api.PodVolumeBackupPhaseFailed) {
+						existPVB.Status.Phase == velerov1api.PodVolumeBackupPhaseFailed ||
+						pvb.Status.Phase == velerov1api.PodVolumeBackupPhaseCanceled) {
 						statusChangedToFinal = false
 					}
 				}
@@ -428,7 +430,7 @@ func (b *backupper) WaitAllPodVolumesProcessed(log logrus.FieldLogger) []*velero
 				continue
 			}
 			podVolumeBackups = append(podVolumeBackups, pvb)
-			if pvb.Status.Phase == velerov1api.PodVolumeBackupPhaseFailed {
+			if pvb.Status.Phase == velerov1api.PodVolumeBackupPhaseFailed || pvb.Status.Phase == velerov1api.PodVolumeBackupPhaseCanceled {
 				log.Errorf("pod volume backup failed: %s", pvb.Status.Message)
 			}
 		}