refactor: Optimize VSC handle readiness polling for VSS backups

changelogs/unreleased/9602-sseago

@@ -0,0 +1 @@
+Optimize VSC handle readiness polling for VSS backups

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

@@ -41,6 +41,12 @@ spec:
           spec:
             description: BackupSpec defines the specification for a Velero backup.
             properties:
+              csiSnapshotEarlyFrequentPolling:
+                description: |-
+                  CSISnapshotEarlyFrequentPolling specifies whether to use an early frequent polling for CSI VolumeSnapshot status turns to
+                  ReadyToUse during creation.
+                nullable: true
+                type: boolean
               csiSnapshotTimeout:
                 description: |-
                   CSISnapshotTimeout specifies the time used to wait for CSI VolumeSnapshot status turns to

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

@@ -80,6 +80,12 @@ spec:
                   Template is the definition of the Backup to be run
                   on the provided schedule
                 properties:
+                  csiSnapshotEarlyFrequentPolling:
+                    description: |-
+                      CSISnapshotEarlyFrequentPolling specifies whether to use an early frequent polling for CSI VolumeSnapshot status turns to
+                      ReadyToUse during creation.
+                    nullable: true
+                    type: boolean
                   csiSnapshotTimeout:
                     description: |-
                       CSISnapshotTimeout specifies the time used to wait for CSI VolumeSnapshot status turns to

pkg/apis/velero/v1/backup_types.go

@@ -162,6 +162,12 @@ type BackupSpec struct {
 	// +optional
 	CSISnapshotTimeout metav1.Duration `json:"csiSnapshotTimeout,omitempty"`
 
+	// CSISnapshotEarlyFrequentPolling specifies whether to use an early frequent polling for CSI VolumeSnapshot status turns to
+	// ReadyToUse during creation.
+	// +optional
+	// +nullable
+	CSISnapshotEarlyFrequentPolling *bool `json:"csiSnapshotEarlyFrequentPolling,omitempty"`
+
 	// ItemOperationTimeout specifies the time used to wait for asynchronous BackupItemAction operations
 	// The default value is 4 hour.
 	// +optional

pkg/backup/actions/csi/pvc_action.go

@@ -351,6 +351,7 @@ func (p *pvcBackupItemAction) Execute(
 		p.crClient,
 		p.log,
 		backup.Spec.CSISnapshotTimeout.Duration,
+		backup.Spec.CSISnapshotEarlyFrequentPolling,
 	)
 	if err != nil {
 		p.log.Errorf("Failed to wait for VolumeSnapshot %s/%s to become ReadyToUse within timeout %v: %s",

pkg/builder/backup_builder.go

@@ -288,6 +288,12 @@ func (b *BackupBuilder) CSISnapshotTimeout(timeout time.Duration) *BackupBuilder
 	return b
 }
 
+// CSISnapshotEarlyFrequentPolling sets the Backup's CSISnapshotEarlyFrequentPolling flag.
+func (b *BackupBuilder) CSISnapshotEarlyFrequentPolling(val bool) *BackupBuilder {
+	b.object.Spec.CSISnapshotEarlyFrequentPolling = &val
+	return b
+}
+
 // ItemOperationTimeout sets the Backup's ItemOperationTimeout
 func (b *BackupBuilder) ItemOperationTimeout(timeout time.Duration) *BackupBuilder {
 	b.object.Spec.ItemOperationTimeout.Duration = timeout

pkg/cmd/cli/backup/create.go

@@ -85,6 +85,7 @@ type CreateOptions struct {
 	SnapshotMoveData                flag.OptionalBool
 	DataMover                       string
 	DefaultVolumesToFsBackup        flag.OptionalBool
+	CSISnapshotEarlyFrequentPolling flag.OptionalBool
 	IncludeNamespaces               flag.StringArray
 	ExcludeNamespaces               flag.StringArray
 	IncludeResources                flag.StringArray
@@ -112,11 +113,12 @@ type CreateOptions struct {
 
 func NewCreateOptions() *CreateOptions {
 	return &CreateOptions{
-		IncludeNamespaces:       flag.NewStringArray("*"),
-		Labels:                  flag.NewMap(),
-		Annotations:             flag.NewMap(),
-		SnapshotVolumes:         flag.NewOptionalBool(nil),
-		IncludeClusterResources: flag.NewOptionalBool(nil),
+		IncludeNamespaces:               flag.NewStringArray("*"),
+		Labels:                          flag.NewMap(),
+		Annotations:                     flag.NewMap(),
+		SnapshotVolumes:                 flag.NewOptionalBool(nil),
+		CSISnapshotEarlyFrequentPolling: flag.NewOptionalBool(nil),
+		IncludeClusterResources:         flag.NewOptionalBool(nil),
 	}
 }
 
@@ -153,6 +155,9 @@ func (o *CreateOptions) BindFlags(flags *pflag.FlagSet) {
 	f = flags.VarPF(&o.DefaultVolumesToFsBackup, "default-volumes-to-fs-backup", "", "Use pod volume file system backup by default for volumes")
 	f.NoOptDefVal = cmd.TRUE
 
+	f = flags.VarPF(&o.CSISnapshotEarlyFrequentPolling, "csi-snapshot-early-frequent-polling", "", "Use early frequent polling for CSI snapshot creation by default for all CSI snapshots.")
+	f.NoOptDefVal = cmd.TRUE
+
 	flags.StringVar(&o.ResPoliciesConfigmap, "resource-policies-configmap", "", "Reference to the resource policies configmap that backup should use")
 	flags.StringVar(&o.DataMover, "data-mover", "", "Specify the data mover to be used by the backup. If the parameter is not set or set as 'velero', the built-in data mover will be used")
 	flags.IntVar(&o.ParallelFilesUpload, "parallel-files-upload", 0, "Number of files uploads simultaneously when running a backup. This is only applicable for the kopia uploader")
@@ -414,6 +419,9 @@ func (o *CreateOptions) BuildBackup(namespace string) (*velerov1api.Backup, erro
 		if o.DefaultVolumesToFsBackup.Value != nil {
 			backupBuilder.DefaultVolumesToFsBackup(*o.DefaultVolumesToFsBackup.Value)
 		}
+		if o.CSISnapshotEarlyFrequentPolling.Value != nil {
+			backupBuilder.CSISnapshotEarlyFrequentPolling(*o.CSISnapshotEarlyFrequentPolling.Value)
+		}
 		if o.ResPoliciesConfigmap != "" {
 			backupBuilder.ResourcePolicies(o.ResPoliciesConfigmap)
 		}

pkg/cmd/cli/backup/create_test.go

@@ -39,6 +39,7 @@ import (
 	cmdtest "github.com/vmware-tanzu/velero/pkg/cmd/test"
 	"github.com/vmware-tanzu/velero/pkg/test"
 	velerotest "github.com/vmware-tanzu/velero/pkg/test"
+	"github.com/vmware-tanzu/velero/pkg/util/boolptr"
 )
 
 func TestCreateOptions_BuildBackup(t *testing.T) {
@@ -48,6 +49,7 @@ func TestCreateOptions_BuildBackup(t *testing.T) {
 	o.OrderedResources = "pods=p1,p2;persistentvolumeclaims=pvc1,pvc2"
 	orders, err := ParseOrderedResources(o.OrderedResources)
 	o.CSISnapshotTimeout = 20 * time.Minute
+	o.CSISnapshotEarlyFrequentPolling.Value = boolptr.True()
 	o.ItemOperationTimeout = 20 * time.Minute
 	orLabelSelectors := []*metav1.LabelSelector{
 		{
@@ -64,14 +66,15 @@ func TestCreateOptions_BuildBackup(t *testing.T) {
 	require.NoError(t, err)
 
 	assert.Equal(t, velerov1api.BackupSpec{
-		TTL:                     metav1.Duration{Duration: o.TTL},
-		IncludedNamespaces:      []string(o.IncludeNamespaces),
-		SnapshotVolumes:         o.SnapshotVolumes.Value,
-		IncludeClusterResources: o.IncludeClusterResources.Value,
-		OrderedResources:        orders,
-		OrLabelSelectors:        orLabelSelectors,
-		CSISnapshotTimeout:      metav1.Duration{Duration: o.CSISnapshotTimeout},
-		ItemOperationTimeout:    metav1.Duration{Duration: o.ItemOperationTimeout},
+		TTL:                             metav1.Duration{Duration: o.TTL},
+		IncludedNamespaces:              []string(o.IncludeNamespaces),
+		SnapshotVolumes:                 o.SnapshotVolumes.Value,
+		IncludeClusterResources:         o.IncludeClusterResources.Value,
+		OrderedResources:                orders,
+		OrLabelSelectors:                orLabelSelectors,
+		CSISnapshotTimeout:              metav1.Duration{Duration: o.CSISnapshotTimeout},
+		CSISnapshotEarlyFrequentPolling: o.CSISnapshotEarlyFrequentPolling.Value,
+		ItemOperationTimeout:            metav1.Duration{Duration: o.ItemOperationTimeout},
 	}, backup.Spec)
 
 	assert.Equal(t, map[string]string{
@@ -223,6 +226,7 @@ func TestCreateCommand(t *testing.T) {
 		selector := "a=pod"
 		orderedResources := "pod=pod1,pod2,pod3"
 		csiSnapshotTimeout := "8m30s"
+		csiSnapshotEarlyFrequentPolling := "true"
 		itemOperationTimeout := "99h1m6s"
 		snapshotVolumes := "false"
 		snapshotMoveData := "true"
@@ -252,6 +256,7 @@ func TestCreateCommand(t *testing.T) {
 		flags.Parse([]string{"--selector", selector})
 		flags.Parse([]string{"--ordered-resources", orderedResources})
 		flags.Parse([]string{"--csi-snapshot-timeout", csiSnapshotTimeout})
+		flags.Parse([]string{"--csi-snapshot-early-frequent-polling", csiSnapshotEarlyFrequentPolling})
 		flags.Parse([]string{"--item-operation-timeout", itemOperationTimeout})
 		flags.Parse([]string{fmt.Sprintf("--snapshot-volumes=%s", snapshotVolumes)})
 		flags.Parse([]string{fmt.Sprintf("--snapshot-move-data=%s", snapshotMoveData)})
@@ -302,6 +307,7 @@ func TestCreateCommand(t *testing.T) {
 		require.Equal(t, selector, o.Selector.String())
 		require.Equal(t, orderedResources, o.OrderedResources)
 		require.Equal(t, csiSnapshotTimeout, o.CSISnapshotTimeout.String())
+		require.Equal(t, csiSnapshotEarlyFrequentPolling, o.CSISnapshotEarlyFrequentPolling.String())
 		require.Equal(t, itemOperationTimeout, o.ItemOperationTimeout.String())
 		require.Equal(t, snapshotVolumes, o.SnapshotVolumes.String())
 		require.Equal(t, snapshotMoveData, o.SnapshotMoveData.String())

pkg/cmd/cli/install/install.go

@@ -81,6 +81,7 @@ type Options struct {
 	DefaultVolumesToFsBackup        bool
 	UploaderType                    string
 	DefaultSnapshotMoveData         bool
+	CSISnapshotEarlyFrequentPolling bool
 	DisableInformerCache            bool
 	ScheduleSkipImmediately         bool
 	PodResources                    kubeutil.PodResources
@@ -141,6 +142,7 @@ func (o *Options) BindFlags(flags *pflag.FlagSet) {
 	flags.BoolVar(&o.DefaultVolumesToFsBackup, "default-volumes-to-fs-backup", o.DefaultVolumesToFsBackup, "Bool flag to configure Velero server to use pod volume file system backup by default for all volumes on all backups. Optional.")
 	flags.StringVar(&o.UploaderType, "uploader-type", o.UploaderType, fmt.Sprintf("The type of uploader to transfer the data of pod volumes, supported value: '%s'", uploader.KopiaType))
 	flags.BoolVar(&o.DefaultSnapshotMoveData, "default-snapshot-move-data", o.DefaultSnapshotMoveData, "Bool flag to configure Velero server to move data by default for all snapshots supporting data movement. Optional.")
+	flags.BoolVar(&o.CSISnapshotEarlyFrequentPolling, "csi-snapshot-early-frequent-polling", o.CSISnapshotEarlyFrequentPolling, "Bool flag to configure Velero server to use early frequent polling by default for all CSI snapshots. Optional.")
 	flags.BoolVar(&o.DisableInformerCache, "disable-informer-cache", o.DisableInformerCache, "Disable informer cache for Get calls on restore. With this enabled, it will speed up restore in cases where there are backup resources which already exist in the cluster, but for very large clusters this will increase velero memory usage. Default is false (don't disable). Optional.")
 	flags.BoolVar(&o.ScheduleSkipImmediately, "schedule-skip-immediately", o.ScheduleSkipImmediately, "Skip the first scheduled backup immediately after creating a schedule. Default is false (don't skip).")
 	flags.BoolVar(&o.NodeAgentDisableHostPath, "node-agent-disable-host-path", o.NodeAgentDisableHostPath, "Don't mount the pod volume host path to node-agent. Optional. Pod volume host path mount is required by fs-backup but could be disabled for other backup methods.")
@@ -238,16 +240,17 @@ func NewInstallOptions() *Options {
 		NodeAgentPodCPULimit:      install.DefaultNodeAgentPodCPULimit,
 		NodeAgentPodMemLimit:      install.DefaultNodeAgentPodMemLimit,
 		// Default to creating a VSL unless we're told otherwise
-		UseVolumeSnapshots:       true,
-		NoDefaultBackupLocation:  false,
-		CRDsOnly:                 false,
-		DefaultVolumesToFsBackup: false,
-		UploaderType:             uploader.KopiaType,
-		DefaultSnapshotMoveData:  false,
-		DisableInformerCache:     false,
-		ScheduleSkipImmediately:  false,
-		kubeletRootDir:           install.DefaultKubeletRootDir,
-		NodeAgentDisableHostPath: false,
+		UseVolumeSnapshots:              true,
+		NoDefaultBackupLocation:         false,
+		CRDsOnly:                        false,
+		DefaultVolumesToFsBackup:        false,
+		UploaderType:                    uploader.KopiaType,
+		DefaultSnapshotMoveData:         false,
+		CSISnapshotEarlyFrequentPolling: false,
+		DisableInformerCache:            false,
+		ScheduleSkipImmediately:         false,
+		kubeletRootDir:                  install.DefaultKubeletRootDir,
+		NodeAgentDisableHostPath:        false,
 	}
 }
 
@@ -324,6 +327,7 @@ func (o *Options) AsVeleroOptions() (*install.VeleroOptions, error) {
 		DefaultVolumesToFsBackup:        o.DefaultVolumesToFsBackup,
 		UploaderType:                    o.UploaderType,
 		DefaultSnapshotMoveData:         o.DefaultSnapshotMoveData,
+		CSISnapshotEarlyFrequentPolling: o.CSISnapshotEarlyFrequentPolling,
 		DisableInformerCache:            o.DisableInformerCache,
 		ScheduleSkipImmediately:         o.ScheduleSkipImmediately,
 		PodResources:                    o.PodResources,

pkg/cmd/cli/schedule/create.go

@@ -156,6 +156,7 @@ func (o *CreateOptions) Run(c *cobra.Command, f client.Factory) error {
 				DefaultVolumesToFsBackup:         o.BackupOptions.DefaultVolumesToFsBackup.Value,
 				OrderedResources:                 orders,
 				CSISnapshotTimeout:               metav1.Duration{Duration: o.BackupOptions.CSISnapshotTimeout},
+				CSISnapshotEarlyFrequentPolling:  o.BackupOptions.CSISnapshotEarlyFrequentPolling.Value,
 				ItemOperationTimeout:             metav1.Duration{Duration: o.BackupOptions.ItemOperationTimeout},
 				DataMover:                        o.BackupOptions.DataMover,
 				SnapshotMoveData:                 o.BackupOptions.SnapshotMoveData.Value,