fix(recurringjob): keep snapshots for the backup

Add a new global setting that allows users to keep the retain
number of snapshots when the recurring job does a backup.

Ref: 2997

Signed-off-by: James Lu <james.lu@suse.com>

app/recurring_job.go

@@ -565,11 +565,18 @@ func (job *Job) filterExpiredSnapshotsOfCurrentRecurringJob(snapshotCRs []longho
 	// Only consider deleting the snapshots that were created by our current job
 	snapshotCRs = filterSnapshotCRsWithLabel(snapshotCRs, types.RecurringJobLabel, jobLabel)
 
-	if job.task == longhorn.RecurringJobTypeSnapshot || job.task == longhorn.RecurringJobTypeSnapshotForceCreate {
+	allowBackupSnapshotDeleted, err := job.GetSettingAsBool(types.SettingNameAutoCleanupRecurringJobBackupSnapshot)
+	if err != nil {
+		job.logger.WithError(err).Warnf("Failed to get the setting %v", types.SettingNameAutoCleanupRecurringJobBackupSnapshot)
+		return []string{}
+	}
+
+	// For recurring snapshot job and AutoCleanupRecurringJobBackupSnapshot is disabled, keeps the number of the snapshots as job.retain.
+	if job.task == longhorn.RecurringJobTypeSnapshot || job.task == longhorn.RecurringJobTypeSnapshotForceCreate || !allowBackupSnapshotDeleted {
 		return filterExpiredItems(snapshotCRsToNameWithTimestamps(snapshotCRs), job.retain)
 	}
 
-	// For the recurring backup job, only keep the snapshot of the last backup and the current snapshot
+	// For the recurring backup job, only keep the snapshot of the last backup and the current snapshot when AutoCleanupRecurringJobBackupSnapshot is enabled.
 	retainingSnapshotCRs := map[string]struct{}{job.snapshotName: {}}
 	if !backupDone {
 		lastBackup, err := job.getLastBackup()
@@ -785,6 +792,20 @@ func (job *Job) UpdateVolumeStatus(v *longhorn.Volume) (*longhorn.Volume, error)
 	return job.lhClient.LonghornV1beta2().Volumes(job.namespace).UpdateStatus(context.TODO(), v, metav1.UpdateOptions{})
 }
 
+// GetSettingAsBool returns boolean of the setting value searching by name.
+func (job *Job) GetSettingAsBool(name types.SettingName) (bool, error) {
+	obj, err := job.lhClient.LonghornV1beta2().Settings(job.namespace).Get(context.TODO(), string(name), metav1.GetOptions{})
+	if err != nil {
+		return false, err
+	}
+	value, err := strconv.ParseBool(obj.Value)
+	if err != nil {
+		return false, err
+	}
+
+	return value, nil
+}
+
 // waitForVolumeState timeout in second
 func (job *Job) waitForVolumeState(state string, timeout int) (*longhornclient.Volume, error) {
 	volumeAPI := job.api.Volume

types/setting.go

@@ -82,6 +82,7 @@ const (
 	SettingNameSystemManagedPodsImagePullPolicy                         = SettingName("system-managed-pods-image-pull-policy")
 	SettingNameAllowVolumeCreationWithDegradedAvailability              = SettingName("allow-volume-creation-with-degraded-availability")
 	SettingNameAutoCleanupSystemGeneratedSnapshot                       = SettingName("auto-cleanup-system-generated-snapshot")
+	SettingNameAutoCleanupRecurringJobBackupSnapshot                    = SettingName("auto-cleanup-recurring-job-backup-snapshot")
 	SettingNameConcurrentAutomaticEngineUpgradePerNodeLimit             = SettingName("concurrent-automatic-engine-upgrade-per-node-limit")
 	SettingNameBackingImageCleanupWaitInterval                          = SettingName("backing-image-cleanup-wait-interval")
 	SettingNameBackingImageRecoveryWaitInterval                         = SettingName("backing-image-recovery-wait-interval")
@@ -157,6 +158,7 @@ var (
 		SettingNameSystemManagedPodsImagePullPolicy,
 		SettingNameAllowVolumeCreationWithDegradedAvailability,
 		SettingNameAutoCleanupSystemGeneratedSnapshot,
+		SettingNameAutoCleanupRecurringJobBackupSnapshot,
 		SettingNameConcurrentAutomaticEngineUpgradePerNodeLimit,
 		SettingNameBackingImageCleanupWaitInterval,
 		SettingNameBackingImageRecoveryWaitInterval,
@@ -258,6 +260,7 @@ var (
 		SettingNameSystemManagedPodsImagePullPolicy:                         SettingDefinitionSystemManagedPodsImagePullPolicy,
 		SettingNameAllowVolumeCreationWithDegradedAvailability:              SettingDefinitionAllowVolumeCreationWithDegradedAvailability,
 		SettingNameAutoCleanupSystemGeneratedSnapshot:                       SettingDefinitionAutoCleanupSystemGeneratedSnapshot,
+		SettingNameAutoCleanupRecurringJobBackupSnapshot:                    SettingDefinitionAutoCleanupRecurringJobBackupSnapshot,
 		SettingNameConcurrentAutomaticEngineUpgradePerNodeLimit:             SettingDefinitionConcurrentAutomaticEngineUpgradePerNodeLimit,
 		SettingNameBackingImageCleanupWaitInterval:                          SettingDefinitionBackingImageCleanupWaitInterval,
 		SettingNameBackingImageRecoveryWaitInterval:                         SettingDefinitionBackingImageRecoveryWaitInterval,
@@ -805,6 +808,16 @@ var (
 		Default:     "true",
 	}
 
+	SettingDefinitionAutoCleanupRecurringJobBackupSnapshot = SettingDefinition{
+		DisplayName: "Automatically Cleanup Recurring Job Backup Snapshot",
+		Description: "This setting enables Longhorn to automatically cleanup the snapshot generated by a recurring backup job.",
+		Category:    SettingCategorySnapshot,
+		Type:        SettingTypeBool,
+		Required:    true,
+		ReadOnly:    false,
+		Default:     "true",
+	}
+
 	SettingDefinitionConcurrentAutomaticEngineUpgradePerNodeLimit = SettingDefinition{
 		DisplayName: "Concurrent Automatic Engine Upgrade Per Node Limit",
 		Description: "This setting controls how Longhorn automatically upgrades volumes' engines after upgrading Longhorn manager. " +
@@ -1231,6 +1244,8 @@ func ValidateSetting(name, value string) (err error) {
 		fallthrough
 	case SettingNameAutoCleanupSystemGeneratedSnapshot:
 		fallthrough
+	case SettingNameAutoCleanupRecurringJobBackupSnapshot:
+		fallthrough
 	case SettingNameAutoDeletePodWhenVolumeDetachedUnexpectedly:
 		fallthrough
 	case SettingNameKubernetesClusterAutoscalerEnabled: