if err != nil {
		return err
	}
	wantDirs, err := xl.getDataDirs()
	if err != nil {
		return err
	}

	// Delete all directories we expect to be there.
	for _, dir := range wantDirs {
		delete(foundDirs, dir)
	}

	// Delete excessive directories.
	// Do not abort on context errors.
	for dir := range foundDirs {
		toRemove := pathJoin(volumeDir, path, dir+SlashSeparator)

	K8sServiceHost string
	// KUBERNETES_SERVICE_PORT
	K8sServicePort string
	// KUBERNETES_NODE_NAME
	K8sNodeName string

	// Directory from where the CNI binaries should be copied
	CNIBinSourceDir string
	// Directories into which to copy the CNI binaries
	CNIBinTargetDirs []string

	// The HTTP port for monitoring
	MonitoringPort int

	// The UDS server address that CNI plugin will send log to.
	LogUDSAddress string

		var hint string
		if endpoint.URL != nil {
			hint = fmt.Sprintf("Drive '%s' is not a directory, MinIO erasure coding needs a directory", endpoint.Path)
		} else {
			hint = "Drives are not directories, MinIO erasure coding needs directories"
		}
		logger.Fatal(config.ErrUnableToWriteInBackend(err).Hint(hint), "Unable to initialize backend")
	case errors.Is(err, errDiskAccessDenied):

		}
		select {
		case <-ctx.Done():
			return ctx.Err()
		case resp <- file:
		}
	}
}

// CleanAbandonedData will read metadata of the object on disk
// and delete any data directories and inline data that isn't referenced in metadata.
func (client *storageRESTClient) CleanAbandonedData(ctx context.Context, volume string, path string) error {
	values := make(url.Values)
	values.Set(storageRESTVolume, volume)

		xioutil.SafeClose(resp)
		return err
	}
	defer done(&err)

	return p.storage.ReadMultiple(ctx, req, resp)
}

// CleanAbandonedData will read metadata of the object on disk
// and delete any data directories and inline data that isn't referenced in metadata.
func (p *xlStorageDiskIDCheck) CleanAbandonedData(ctx context.Context, volume string, path string) (err error) {

        `tf.keras.layers.EinsumDense`.
    *   Added `tf.keras.utils.audio_dataset_from_directory` utility to easily
        generate audio classification datasets from directories of `.wav` files.
    *   Added `subset="both"` support in
        `tf.keras.utils.image_dataset_from_directory`,`tf.keras.utils.text_dataset_from_directory`,
        and `audio_dataset_from_directory`, to be used with the