headers: []*Header{{
			Name:       "test2/",
			Mode:       16877,
			Uid:        1000,
			Gid:        1000,
			Size:       14,
			ModTime:    time.Unix(1441973427, 0),
			Typeflag:   'D',
			Uname:      "rawr",
			Gname:      "dsnet",
			AccessTime: time.Unix(1441974501, 0),
			ChangeTime: time.Unix(1441973436, 0),
			Format:     FormatGNU,
		}, {
			Name:       "test2/foo",
			Mode:       33188,

		tests: []testFnc{
			testHeader{Header{
				Typeflag: TypeReg,
				Name:     "hi\x80\x81\x82\x83bye",
				Mode:     0644,
				Uid:      1000,
				Gid:      1000,
				Uname:    "rawr",
				Gname:    "dsnet",
				ModTime:  time.Unix(0, 0),
				Format:   FormatGNU,
			}, nil},
			testClose{nil},
		},
		// TODO(dsnet): Re-enable this test when adding sparse support.

Jinzhu <******@****.***> 1748158840 +0800

Write throughput is typically at least 500MB/s per CPU core,
and scales with the number of available CPU cores.
Decompression speed is typically at least 1GB/s.

This means that in cases where raw IO is below these numbers
compression will not only reduce disk usage but also help increase system throughput.
Typically, enabling compression on spinning disk systems
will increase speed when the content can be compressed.

	case ChecksumCRC32C:
		v := crc32Combine(crc32.Castagnoli, binary.BigEndian.Uint32(c.Raw), binary.BigEndian.Uint32(other.Raw), size)
		binary.BigEndian.PutUint32(c.Raw, v)
	case ChecksumCRC64NVME:
		v := crc64Combine(bits.Reverse64(crc64NVMEPolynomial), binary.BigEndian.Uint64(c.Raw), binary.BigEndian.Uint64(other.Raw), size)
		binary.BigEndian.PutUint64(c.Raw, v)
	default:
		return fmt.Errorf("unknown checksum type: %s", c.Type.String())

Refer to the dashboard [json file here](https://raw.githubusercontent.com/minio/minio/master/docs/metrics/prometheus/grafana/minio-dashboard.json).

![Grafana](https://raw.githubusercontent.com/minio/minio/master/docs/metrics/prometheus/grafana/grafana-minio.png)

import javax.inject.Inject;
import javax.inject.Named;
import javax.inject.Singleton;

/**
 * {@code rar} packaging plugins bindings provider for {@code default} lifecycle.
 */
@Named("rar")
@Singleton
public final class RarLifecycleMappingProvider extends AbstractLifecycleMappingProvider {
    // START SNIPPET: rar
    private static final String[] BINDINGS = {
        "process-resources",

![ambari-login](https://github.com/minio/minio/blob/master/docs/bigdata/images/image3.png?raw=true "ambari login")

### **3.1 Configure Hadoop**

Navigate to **Services** -> **HDFS** -> **CONFIGS** -> **ADVANCED** as shown below

![hdfs-configs](https://github.com/minio/minio/blob/master/docs/bigdata/images/image2.png?raw=true "hdfs advanced configs")

    /**
     * Gets the raw model as it was read from the input model source. Apart from basic validation, the raw model has not
     * undergone any updates by the model builder, e.g. reflects neither inheritance nor interpolation.
     *
     * @return The raw model, never {@code null}.
     */
    Model getRawModel();

    /**

 * <p>
 * Model transformers allow plugins and extensions to modify the POM model during the build process.
 * Transformations can be applied at three different stages:
 * <ol>
 *   <li>File model - The raw model as read directly from the file</li>
 *   <li>Raw model - The model after inheritance has been applied</li>
 *   <li>Effective model - The fully processed model with all interpolation and inheritance applied</li>
 * </ol>
 * <p>