Cherry pick the in-place restore implementation PRs from feature branch to main (#10415)

* Update CRDs and CLI to support in-place restore (#10038) Update CRDs(Restore, DataDownload, PodVolumeRestore) and restore create CLI to support in-place restore Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> * Update Kopia(filesystem) uploader to support incremental and deleteExtraFile during restore (#10066) Update Kopia(filesystem) uploader to support incremental and deleteExtraFile during restore Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> * Update Restore Exposer and PVC CSI to support in-place restore (#10104) 1. Update Restore Exposer to support exposing with existing PV for in-place restore 2. Update PVC CSI RIA to continue the restore process for in-place restore Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> * Update Block uploader to support increase restore (#10244) Update Block uploader to support increase restore Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> * Update Exposer to recreate the target PV if the volume mode is different with the restore PVC (#10257) Update Exposer to recreate the target PV if the volume mode is different with t he restore PVC Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> * Preserve PVC selected-node annotation via carrier annotation for in-place restore For in-place volume data restore, the existing PVC is deleted and recreated. For StorageClasses with the WaitForFirstConsumer volume binding mode, losing the volume.kubernetes.io/selected-node annotation could let the scheduler place the recreated workload Pod in a different zone than the original PV, leaving it stuck in ContainerCreating. Instead of relying on RestoreItemAction execution order (the generic PVC RIA unconditionally strips the selected-node annotation), the PVC CSI RIA now captures the annotation from the existing PVC right before deleting it and carries it on the target PVC via the Velero-internal restore.velero.io/inplace-restore-selected-node annotation. The restore engine translates the carrier back to the Kubernetes annotation after all RestoreItemActions have run and always strips the carrier so it never lands on the cluster. This makes the behavior independent of RIA ordering: the Kubernetes annotation is stripped by default on every path (including when the target PVC does not exist and Velero falls back to provisioning a new PVC), and preservation only happens when the CSI RIA explicitly captured a value from the existing PVC. Signed-off-by: chlins <chlins.zhang@gmail.com> * Update the control path to make the in-place incremental restore with block data mover work E2E (#10410) Update the control path to make the in-place incremental restore with block data mover work E2E Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> --------- Signed-off-by: Wenkai Yin(尹文开) <yinw@vmware.com> Signed-off-by: chlins <chlins.zhang@gmail.com> Co-authored-by: chlins <chlins.zhang@gmail.com>
2026-09-13 11:34:54 +00:00 · 2026-08-26 10:21:32 -04:00
parent ac5744c7b4
commit b7d83a6f2b
54 changed files with 2164 additions and 335 deletions
@@ -36,6 +36,7 @@ import (
 	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"
+	"github.com/vmware-tanzu/velero/pkg/cbtservice"
 	"github.com/vmware-tanzu/velero/pkg/client"
 	"github.com/vmware-tanzu/velero/pkg/cmd/util/signals"
 	"github.com/vmware-tanzu/velero/pkg/datamover"
@@ -56,6 +57,9 @@ type dataMoverRestoreConfig struct {
 	ddName          string
 	cacheDir        string
 	resourceTimeout time.Duration
+	cbtSAName       string
+	vsNamespace     string
+	volumeID        string
 }

 func NewRestoreCommand(f client.Factory) *cobra.Command {
@@ -96,6 +100,9 @@ func NewRestoreCommand(f client.Factory) *cobra.Command {
 	command.Flags().StringVar(&config.ddName, "data-download", config.ddName, "The data download name")
 	command.Flags().StringVar(&config.cacheDir, "cache-volume-path", config.cacheDir, "The full path of the cache volume")
 	command.Flags().DurationVar(&config.resourceTimeout, "resource-timeout", config.resourceTimeout, "How long to wait for resource processes which are not covered by other specific timeout parameters.")
+	command.Flags().StringVar(&config.cbtSAName, "cbt-sa-name", config.cbtSAName, "The name of the service account used by CSI's CBT service")
+	command.Flags().StringVar(&config.vsNamespace, "vs-namespace", config.vsNamespace, "The namespace of the VolumeSnapshot")
+	command.Flags().StringVar(&config.volumeID, "volume-id", config.volumeID, "The volume ID of the snapshot")

 	_ = command.MarkFlagRequired("volume-path")
 	_ = command.MarkFlagRequired("volume-mode")
@@ -116,6 +123,7 @@ type dataMoverRestore struct {
 	config      dataMoverRestoreConfig
 	kubeClient  kubernetes.Interface
 	dataPathMgr *datapath.Manager
+	cbtService  cbtservice.Service
 }

 func newdataMoverRestore(logger logrus.FieldLogger, factory client.Factory, config dataMoverRestoreConfig) (*dataMoverRestore, error) {
@@ -201,6 +209,12 @@ func newdataMoverRestore(logger logrus.FieldLogger, factory client.Factory, conf
 		config:     config,
 		namespace:  factory.Namespace(),
 		nodeName:   nodeName,
+		cbtService: cbtservice.NewService(
+			logger,
+			config.vsNamespace,
+			config.cbtSAName,
+			clientConfig,
+		),
 	}

 	s.kubeClient, err = factory.KubeClient()
@@ -294,5 +308,5 @@ func (s *dataMoverRestore) createDataPathService() (dataPathService, error) {
 	return datamover.NewRestoreMicroService(s.ctx, s.client, s.kubeClient, s.config.ddName, s.namespace, s.nodeName, datapath.AccessPoint{
 		ByPath:  s.config.volumePath,
 		VolMode: uploader.PersistentVolumeMode(s.config.volumeMode),
-	}, s.dataPathMgr, repoEnsurer, credGetter, duInformer, s.config.cacheDir, s.logger), nil
+	}, s.dataPathMgr, repoEnsurer, credGetter, duInformer, s.config.cacheDir, s.config.volumeID, s.cbtService, s.logger), nil
 }
@@ -27,6 +27,7 @@ import (

 	"github.com/bombsimon/logrusr/v3"
 	"github.com/cockroachdb/errors"
+	snapshotv1api "github.com/kubernetes-csi/external-snapshotter/client/v8/apis/volumesnapshot/v1"
 	snapshotv1client "github.com/kubernetes-csi/external-snapshotter/client/v8/clientset/versioned"
 	"github.com/prometheus/client_golang/prometheus/promhttp"
 	"github.com/sirupsen/logrus"
@@ -175,6 +176,10 @@ func newNodeAgentServer(logger logrus.FieldLogger, factory client.Factory, confi
 		cancelFunc()
 		return nil, err
 	}
+	if err := snapshotv1api.AddToScheme(scheme); err != nil {
+		cancelFunc()
+		return nil, err
+	}

 	nodeName := os.Getenv("NODE_NAME")

@@ -484,6 +489,7 @@ func (s *nodeAgentServer) run() {
 		s.repoConfigMgr,
 		podLabels,
 		podAnnotations,
+		csiSnapshotMetadataServiceConfigs,
 	)

 	if err := dataDownloadReconciler.SetupWithManager(s.mgr); err != nil {
@@ -99,6 +99,7 @@ type CreateOptions struct {
 	IncludeNamespaces           flag.StringArray
 	ExcludeNamespaces           flag.StringArray
 	ExistingResourcePolicy      string
+	ExistingVolumeDataPolicy    string
 	IncludeResources            flag.StringArray
 	ExcludeResources            flag.StringArray
 	StatusIncludeResources      flag.StringArray
@@ -115,6 +116,7 @@ type CreateOptions struct {
 	SkipDefaultResourceModifier bool
 	WriteSparseFiles            flag.OptionalBool
 	ParallelFilesDownload       int
+	DeleteExtraFiles            flag.OptionalBool
 	client                      kbclient.WithWatch
 }

@@ -128,6 +130,7 @@ func NewCreateOptions() *CreateOptions {
 		PreserveNodePorts:       flag.NewOptionalBool(nil),
 		IncludeClusterResources: flag.NewOptionalBool(nil),
 		WriteSparseFiles:        flag.NewOptionalBool(nil),
+		DeleteExtraFiles:        flag.NewOptionalBool(nil),
 	}
 }

@@ -141,7 +144,8 @@ func (o *CreateOptions) BindFlags(flags *pflag.FlagSet) {
 	flags.Var(&o.Annotations, "annotations", "Annotations to apply to the restore.")
 	flags.Var(&o.IncludeResources, "include-resources", "Resources to include in the restore, formatted as resource.group, such as storageclasses.storage.k8s.io (use '*' for all resources).")
 	flags.Var(&o.ExcludeResources, "exclude-resources", "Resources to exclude from the restore, formatted as resource.group, such as storageclasses.storage.k8s.io.")
-	flags.StringVar(&o.ExistingResourcePolicy, "existing-resource-policy", "", "Restore Policy to be used during the restore workflow, can be - none or update")
+	flags.StringVar(&o.ExistingResourcePolicy, "existing-resource-policy", "", "Restore Policy to be used during the restore workflow for Kubernetes resources, can be - none or update")
+	flags.StringVar(&o.ExistingVolumeDataPolicy, "existing-volume-data-policy", "", "Restore Policy to be used during the restore workflow for volume data, can be - none, full or incremental")
 	flags.Var(&o.StatusIncludeResources, "status-include-resources", "Resources to include in the restore status, formatted as resource.group, such as storageclasses.storage.k8s.io.")
 	flags.Var(&o.StatusExcludeResources, "status-exclude-resources", "Resources to exclude from the restore status, formatted as resource.group, such as storageclasses.storage.k8s.io.")
 	flags.VarP(&o.Selector, "selector", "l", "Only restore resources matching this label selector.")
@@ -175,6 +179,9 @@ func (o *CreateOptions) BindFlags(flags *pflag.FlagSet) {
 	f.NoOptDefVal = cmd.TRUE

 	flags.IntVar(&o.ParallelFilesDownload, "parallel-files-download", 0, "The number of restore operations to run in parallel. If set to 0, the default parallelism will be the number of CPUs for the node that node agent pod is running.")
+
+	f = flags.VarPF(&o.DeleteExtraFiles, "delete-extra-files", "", "Whether to delete extra files in the target volume that do not exist in the backup during file system restore. This setting is only applicable to File System restores (PodVolumeBackup or CSI File System Data Move) and has no effect on Block Data Move restores.")
+	f.NoOptDefVal = cmd.TRUE
 }

 func (o *CreateOptions) Complete(args []string, f client.Factory) error {
@@ -224,6 +231,10 @@ func (o *CreateOptions) Validate(c *cobra.Command, args []string, f client.Facto
 		return errors.New("existing-resource-policy has invalid value, it accepts only none, update as value")
 	}

+	if len(o.ExistingVolumeDataPolicy) > 0 && !restore.IsVolumeDataPolicyValid(o.ExistingVolumeDataPolicy) {
+		return errors.New("existing-volume-data-policy has invalid value, it accepts only none, full, incremental as value")
+	}
+
 	if o.ParallelFilesDownload < 0 {
 		return errors.New("parallel-files-download cannot be negative")
 	}
@@ -344,27 +355,29 @@ func (o *CreateOptions) Run(c *cobra.Command, f client.Factory) error {
 			Annotations: o.Annotations.Data(),
 		},
 		Spec: api.RestoreSpec{
-			BackupName:              o.BackupName,
-			ScheduleName:            o.ScheduleName,
-			IncludedNamespaces:      o.IncludeNamespaces,
-			ExcludedNamespaces:      o.ExcludeNamespaces,
-			IncludedResources:       o.IncludeResources,
-			ExcludedResources:       o.ExcludeResources,
-			ExistingResourcePolicy:  api.PolicyType(o.ExistingResourcePolicy),
-			NamespaceMapping:        o.NamespaceMappings.Data(),
-			LabelSelector:           o.Selector.LabelSelector,
-			OrLabelSelectors:        o.OrSelector.OrLabelSelectors,
-			RestorePVs:              o.RestoreVolumes.Value,
-			PreserveNodePorts:       o.PreserveNodePorts.Value,
-			IncludeClusterResources: o.IncludeClusterResources.Value,
-			ResourceModifier:        resModifiers,
-			ResourcePolicy:          resPolicies,
+			BackupName:               o.BackupName,
+			ScheduleName:             o.ScheduleName,
+			IncludedNamespaces:       o.IncludeNamespaces,
+			ExcludedNamespaces:       o.ExcludeNamespaces,
+			IncludedResources:        o.IncludeResources,
+			ExcludedResources:        o.ExcludeResources,
+			ExistingResourcePolicy:   api.ResourcePolicyType(o.ExistingResourcePolicy),
+			ExistingVolumeDataPolicy: api.VolumeDataPolicyType(o.ExistingVolumeDataPolicy),
+			NamespaceMapping:         o.NamespaceMappings.Data(),
+			LabelSelector:            o.Selector.LabelSelector,
+			OrLabelSelectors:         o.OrSelector.OrLabelSelectors,
+			RestorePVs:               o.RestoreVolumes.Value,
+			PreserveNodePorts:        o.PreserveNodePorts.Value,
+			IncludeClusterResources:  o.IncludeClusterResources.Value,
+			ResourceModifier:         resModifiers,
+			ResourcePolicy:           resPolicies,
 			ItemOperationTimeout: metav1.Duration{
 				Duration: o.ItemOperationTimeout,
 			},
 			UploaderConfig: &api.UploaderConfigForRestore{
 				WriteSparseFiles:      o.WriteSparseFiles.Value,
 				ParallelFilesDownload: o.ParallelFilesDownload,
+				DeleteExtraFiles:      o.DeleteExtraFiles.Value,
 			},
 		},
 	}
@@ -68,6 +68,7 @@ func TestCreateCommand(t *testing.T) {
 		includeNamespaces := "app1,app2"
 		excludeNamespaces := "pod1,pod2,pod3"
 		existingResourcePolicy := "none"
+		existingVolumeDataPolicy := "none"
 		includeResources := "sc,sts"
 		excludeResources := "job"
 		statusIncludeResources := "sc,sts"
@@ -80,6 +81,7 @@ func TestCreateCommand(t *testing.T) {
 		resourceModifierConfigMap := "modifier-cm"
 		ResourcePoliciesConfigMap := "policies-cm"
 		writeSparseFiles := "true"
+		deleteExtraFiles := "true"
 		parallel := 2
 		flags := new(pflag.FlagSet)
 		o := NewCreateOptions()
@@ -92,6 +94,7 @@ func TestCreateCommand(t *testing.T) {
 		flags.Parse([]string{"--labels", labels})
 		flags.Parse([]string{"--annotations", annotations})
 		flags.Parse([]string{"--existing-resource-policy", existingResourcePolicy})
+		flags.Parse([]string{"--existing-volume-data-policy", existingVolumeDataPolicy})
 		flags.Parse([]string{"--include-namespaces", includeNamespaces})
 		flags.Parse([]string{"--exclude-namespaces", excludeNamespaces})
 		flags.Parse([]string{"--include-resources", includeResources})
@@ -107,6 +110,7 @@ func TestCreateCommand(t *testing.T) {
 		flags.Parse([]string{"--resource-policies-configmap", ResourcePoliciesConfigMap})
 		flags.Parse([]string{"--skip-default-resource-modifier"})
 		flags.Parse([]string{"--write-sparse-files", writeSparseFiles})
+		flags.Parse([]string{"--delete-extra-files", deleteExtraFiles})
 		flags.Parse([]string{"--parallel-files-download", "2"})
 		client := velerotest.NewFakeControllerRuntimeClient(t).(kbclient.WithWatch)

@@ -134,6 +138,7 @@ func TestCreateCommand(t *testing.T) {
 		require.Equal(t, includeNamespaces, o.IncludeNamespaces.String())
 		require.Equal(t, excludeNamespaces, o.ExcludeNamespaces.String())
 		require.Equal(t, existingResourcePolicy, o.ExistingResourcePolicy)
+		require.Equal(t, existingVolumeDataPolicy, o.ExistingVolumeDataPolicy)
 		require.Equal(t, includeResources, o.IncludeResources.String())
 		require.Equal(t, excludeResources, o.ExcludeResources.String())

@@ -149,6 +154,7 @@ func TestCreateCommand(t *testing.T) {
 		require.True(t, o.SkipDefaultResourceModifier)
 		require.Equal(t, writeSparseFiles, o.WriteSparseFiles.String())
 		require.Equal(t, parallel, o.ParallelFilesDownload)
+		require.Equal(t, deleteExtraFiles, o.DeleteExtraFiles.String())
 	})

 	t.Run("create a restore from schedule", func(t *testing.T) {