Java FileSystem.getFileChecksum方法代码示例

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Override
public PathMetadata apply(@Nonnull Path location) {
  try {
    FileSystem fs = location.getFileSystem(conf);
    FileStatus fileStatus = fs.getFileStatus(location);
    FileChecksum checksum = null;
    if (fileStatus.isFile()) {
      checksum = fs.getFileChecksum(location);
    }

    List<PathMetadata> childPathDescriptors = new ArrayList<>();
    if (fileStatus.isDirectory()) {
      FileStatus[] childStatuses = fs.listStatus(location);
      for (FileStatus childStatus : childStatuses) {
        childPathDescriptors.add(apply(childStatus.getPath()));
      }
    }

    return new PathMetadata(location, fileStatus.getModificationTime(), checksum, childPathDescriptors);

  } catch (IOException e) {
    throw new CircusTrainException("Unable to compute digest for location " + location.toString(), e);
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private void testChecksum() throws Exception {
  if (!isLocalFS()) {
    FileSystem fs = FileSystem.get(getProxiedFSConf());
    fs.mkdirs(getProxiedFSTestDir());
    Path path = new Path(getProxiedFSTestDir(), "foo.txt");
    OutputStream os = fs.create(path);
    os.write(1);
    os.close();
    FileChecksum hdfsChecksum = fs.getFileChecksum(path);
    fs.close();
    fs = getHttpFSFileSystem();
    FileChecksum httpChecksum = fs.getFileChecksum(path);
    fs.close();
    Assert.assertEquals(httpChecksum.getAlgorithmName(), hdfsChecksum.getAlgorithmName());
    Assert.assertEquals(httpChecksum.getLength(), hdfsChecksum.getLength());
    Assert.assertArrayEquals(httpChecksum.getBytes(), hdfsChecksum.getBytes());
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
/**
 * 此方法用于获取文件的FileChecksum
 *
 * @param fileSystemInfo
 *            文件系统信息
 * @param path
 *            文件路径
 * @return FileChecksum
 */
public static FileChecksum getFileChecksum(FileSystemInfo fileSystemInfo, String path) {
	FileSystem fs = getFileSystem(fileSystemInfo);
	Path uri = new Path(path);
	try {
		pathNotExistCheck(path, fs, uri);
		return fs.getFileChecksum(uri);
	} catch (IOException e) {
		e.printStackTrace();
	} finally {
		closeFileSystem(fs);
	}
	return null;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testEncryptedReadAfterNameNodeRestart() throws IOException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    cluster = new MiniDFSCluster.Builder(conf).build();
    
    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();
    
    setEncryptionConfigKeys(conf);
    
    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();
    
    fs = getFileSystem(conf);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    fs.close();
    
    cluster.restartNameNode();
    fs = getFileSystem(conf);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    fs.close();
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testGetFileChecksum() throws Exception {
  final String f = "/testGetFileChecksum";
  final Path p = new Path(f);

  final MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).numDataNodes(3).build();
  try {
    cluster.waitActive();

    //create a file
    final FileSystem fs = cluster.getFileSystem();
    DFSTestUtil.createFile(fs, p, 1L << 20, (short)3, 20100402L);

    //get checksum
    final FileChecksum cs1 = fs.getFileChecksum(p);
    assertTrue(cs1 != null);

    //stop the first datanode
    final List<LocatedBlock> locatedblocks = DFSClient.callGetBlockLocations(
        cluster.getNameNodeRpc(), f, 0, Long.MAX_VALUE)
          .getLocatedBlocks();
    final DatanodeInfo first = locatedblocks.get(0).getLocations()[0];
    cluster.stopDataNode(first.getXferAddr());

    //get checksum again
    final FileChecksum cs2 = fs.getFileChecksum(p);
    assertEquals(cs1, cs2);
  } finally {
    cluster.shutdown();
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private FileAction checkUpdate(FileSystem sourceFS, FileStatus source,
    Path target) throws IOException {
  final FileStatus targetFileStatus;
  try {
    targetFileStatus = targetFS.getFileStatus(target);
  } catch (FileNotFoundException e) {
    return FileAction.OVERWRITE;
  }
  if (targetFileStatus != null && !overWrite) {
    if (canSkip(sourceFS, source, targetFileStatus)) {
      return FileAction.SKIP;
    } else if (append) {
      long targetLen = targetFileStatus.getLen();
      if (targetLen < source.getLen()) {
        FileChecksum sourceChecksum = sourceFS.getFileChecksum(
            source.getPath(), targetLen);
        if (sourceChecksum != null
            && sourceChecksum.equals(targetFS.getFileChecksum(target))) {
          // We require that the checksum is not null. Thus currently only
          // DistributedFileSystem is supported
          return FileAction.APPEND;
        }
      }
    }
  }
  return FileAction.OVERWRITE;
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private FileChecksum getFileChecksum(final FileSystem fs, final Path path) {
  try {
    return fs.getFileChecksum(path);
  } catch (IOException e) {
    LOG.warn("Unable to get checksum for file=" + path, e);
    return null;
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testCreateWithCustomChecksum() throws Exception {
  Configuration conf = getTestConfiguration();
  MiniDFSCluster cluster = null;
  Path testBasePath = new Path("/test/csum");
  // create args 
  Path path1 = new Path(testBasePath, "file_wtih_crc1");
  Path path2 = new Path(testBasePath, "file_with_crc2");
  ChecksumOpt opt1 = new ChecksumOpt(DataChecksum.Type.CRC32C, 512);
  ChecksumOpt opt2 = new ChecksumOpt(DataChecksum.Type.CRC32, 512);

  // common args
  FsPermission perm = FsPermission.getDefault().applyUMask(
      FsPermission.getUMask(conf));
  EnumSet<CreateFlag> flags = EnumSet.of(CreateFlag.OVERWRITE,
      CreateFlag.CREATE);
  short repl = 1;

  try {
    cluster = new MiniDFSCluster.Builder(conf).numDataNodes(1).build();
    FileSystem dfs = cluster.getFileSystem();

    dfs.mkdirs(testBasePath);

    // create two files with different checksum types
    FSDataOutputStream out1 = dfs.create(path1, perm, flags, 4096, repl,
        131072L, null, opt1);
    FSDataOutputStream out2 = dfs.create(path2, perm, flags, 4096, repl,
        131072L, null, opt2);

    for (int i = 0; i < 1024; i++) {
      out1.write(i);
      out2.write(i);
    }
    out1.close();
    out2.close();

    // the two checksums must be different.
    MD5MD5CRC32FileChecksum sum1 =
        (MD5MD5CRC32FileChecksum)dfs.getFileChecksum(path1);
    MD5MD5CRC32FileChecksum sum2 =
        (MD5MD5CRC32FileChecksum)dfs.getFileChecksum(path2);
    assertFalse(sum1.equals(sum2));

    // check the individual params
    assertEquals(DataChecksum.Type.CRC32C, sum1.getCrcType());
    assertEquals(DataChecksum.Type.CRC32,  sum2.getCrcType());

  } finally {
    if (cluster != null) {
      cluster.getFileSystem().delete(testBasePath, true);
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testEncryptedRead() throws IOException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    cluster = new MiniDFSCluster.Builder(conf).build();
    
    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();
    
    setEncryptionConfigKeys(conf);
    
    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();
    
    fs = getFileSystem(conf);
    LogCapturer logs = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(SaslDataTransferServer.class));
    LogCapturer logs1 = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(DataTransferSaslUtil.class));
    try {
      assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
      assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    } finally {
      logs.stopCapturing();
      logs1.stopCapturing();
    }
    
    fs.close();
    
    if (resolverClazz == null) {
      // Test client and server negotiate cipher option
      GenericTestUtils.assertDoesNotMatch(logs.getOutput(),
          "Server using cipher suite");
      // Check the IOStreamPair
      GenericTestUtils.assertDoesNotMatch(logs1.getOutput(),
          "Creating IOStreamPair of CryptoInputStream and CryptoOutputStream.");
    }
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testEncryptedReadWithRC4() throws IOException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    cluster = new MiniDFSCluster.Builder(conf).build();
    
    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();
    
    setEncryptionConfigKeys(conf);
    // It'll use 3DES by default, but we set it to rc4 here.
    conf.set(DFSConfigKeys.DFS_DATA_ENCRYPTION_ALGORITHM_KEY, "rc4");
    
    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();
    
    fs = getFileSystem(conf);
    LogCapturer logs = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(SaslDataTransferServer.class));
    LogCapturer logs1 = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(DataTransferSaslUtil.class));
    try {
      assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
      assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    } finally {
      logs.stopCapturing();
      logs1.stopCapturing();
    }

    fs.close();

    if (resolverClazz == null) {
      // Test client and server negotiate cipher option
      GenericTestUtils.assertDoesNotMatch(logs.getOutput(),
          "Server using cipher suite");
      // Check the IOStreamPair
      GenericTestUtils.assertDoesNotMatch(logs1.getOutput(),
          "Creating IOStreamPair of CryptoInputStream and CryptoOutputStream.");
    }
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testEncryptedReadWithAES() throws IOException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    conf.set(DFSConfigKeys.DFS_ENCRYPT_DATA_TRANSFER_CIPHER_SUITES_KEY,
        "AES/CTR/NoPadding");
    cluster = new MiniDFSCluster.Builder(conf).build();

    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();

    setEncryptionConfigKeys(conf);

    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();

    fs = getFileSystem(conf);
    LogCapturer logs = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(SaslDataTransferServer.class));
    LogCapturer logs1 = GenericTestUtils.LogCapturer.captureLogs(
        LogFactory.getLog(DataTransferSaslUtil.class));
    try {
      assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
      assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    } finally {
      logs.stopCapturing();
      logs1.stopCapturing();
    }

    fs.close();

    if (resolverClazz == null) {
      // Test client and server negotiate cipher option
      GenericTestUtils.assertMatches(logs.getOutput(),
          "Server using cipher suite");
      // Check the IOStreamPair
      GenericTestUtils.assertMatches(logs1.getOutput(),
          "Creating IOStreamPair of CryptoInputStream and CryptoOutputStream.");
    }
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testLongLivedReadClientAfterRestart() throws IOException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    cluster = new MiniDFSCluster.Builder(conf).build();
    
    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();
    
    setEncryptionConfigKeys(conf);
    
    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();
    
    fs = getFileSystem(conf);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    
    // Restart the NN and DN, after which the client's encryption key will no
    // longer be valid.
    cluster.restartNameNode();
    assertTrue(cluster.restartDataNode(0));
    
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    
    fs.close();
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
@Test
public void testLongLivedClient() throws IOException, InterruptedException {
  MiniDFSCluster cluster = null;
  try {
    Configuration conf = new Configuration();
    cluster = new MiniDFSCluster.Builder(conf).build();
    
    FileSystem fs = getFileSystem(conf);
    writeTestDataToFile(fs);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    FileChecksum checksum = fs.getFileChecksum(TEST_PATH);
    fs.close();
    cluster.shutdown();
    
    setEncryptionConfigKeys(conf);
    
    cluster = new MiniDFSCluster.Builder(conf)
        .manageDataDfsDirs(false)
        .manageNameDfsDirs(false)
        .format(false)
        .startupOption(StartupOption.REGULAR)
        .build();
    
    BlockTokenSecretManager btsm = cluster.getNamesystem().getBlockManager()
        .getBlockTokenSecretManager();
    btsm.setKeyUpdateIntervalForTesting(2 * 1000);
    btsm.setTokenLifetime(2 * 1000);
    btsm.clearAllKeysForTesting();
    
    fs = getFileSystem(conf);
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    
    // Sleep for 15 seconds, after which the encryption key will no longer be
    // valid. It needs to be a few multiples of the block token lifetime,
    // since several block tokens are valid at any given time (the current
    // and the last two, by default.)
    LOG.info("Sleeping so that encryption keys expire...");
    Thread.sleep(15 * 1000);
    LOG.info("Done sleeping.");
    
    assertEquals(PLAIN_TEXT, DFSTestUtil.readFile(fs, TEST_PATH));
    assertEquals(checksum, fs.getFileChecksum(TEST_PATH));
    
    fs.close();
  } finally {
    if (cluster != null) {
      cluster.shutdown();
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
private long doCopy(FileStatus sourceFileStatus, Path target,
    Mapper.Context context, EnumSet<FileAttribute> fileAttributes)
    throws IOException {
  final boolean toAppend = action == FileAction.APPEND;
  Path targetPath = toAppend ? target : getTmpFile(target, context);
  final Configuration configuration = context.getConfiguration();
  FileSystem targetFS = target.getFileSystem(configuration);

  try {
    if (LOG.isDebugEnabled()) {
      LOG.debug("Copying " + sourceFileStatus.getPath() + " to " + target);
      LOG.debug("Target file path: " + targetPath);
    }
    final Path sourcePath = sourceFileStatus.getPath();
    final FileSystem sourceFS = sourcePath.getFileSystem(configuration);
    final FileChecksum sourceChecksum = fileAttributes
        .contains(FileAttribute.CHECKSUMTYPE) ? sourceFS
        .getFileChecksum(sourcePath) : null;

    final long offset = action == FileAction.APPEND ? targetFS.getFileStatus(
        target).getLen() : 0;
    long bytesRead = copyToFile(targetPath, targetFS, sourceFileStatus,
        offset, context, fileAttributes, sourceChecksum);

    compareFileLengths(sourceFileStatus, targetPath, configuration, bytesRead
        + offset);
    //At this point, src&dest lengths are same. if length==0, we skip checksum
    if ((bytesRead != 0) && (!skipCrc)) {
      compareCheckSums(sourceFS, sourceFileStatus.getPath(), sourceChecksum,
          targetFS, targetPath);
    }
    // it's not append case, thus we first write to a temporary file, rename
    // it to the target path.
    if (!toAppend) {
      promoteTmpToTarget(targetPath, target, targetFS);
    }
    return bytesRead;
  } finally {
    // note that for append case, it is possible that we append partial data
    // and then fail. In that case, for the next retry, we either reuse the
    // partial appended data if it is good or we overwrite the whole file
    if (!toAppend && targetFS.exists(targetPath)) {
      targetFS.delete(targetPath, false);
    }
  }
}
 

import org.apache.hadoop.fs.FileSystem; //导入方法依赖的package包/类
/**
 * Utility to compare checksums for the paths specified.
 *
 * If checksums's can't be retrieved, it doesn't fail the test
 * Only time the comparison would fail is when checksums are
 * available and they don't match
 *
 * @param sourceFS FileSystem for the source path.
 * @param source The source path.
 * @param sourceChecksum The checksum of the source file. If it is null we
 * still need to retrieve it through sourceFS.
 * @param targetFS FileSystem for the target path.
 * @param target The target path.
 * @return If either checksum couldn't be retrieved, the function returns
 * false. If checksums are retrieved, the function returns true if they match,
 * and false otherwise.
 * @throws IOException if there's an exception while retrieving checksums.
 */
public static boolean checksumsAreEqual(FileSystem sourceFS, Path source,
    FileChecksum sourceChecksum, FileSystem targetFS, Path target)
    throws IOException {
  FileChecksum targetChecksum = null;
  try {
    sourceChecksum = sourceChecksum != null ? sourceChecksum : sourceFS
        .getFileChecksum(source);
    targetChecksum = targetFS.getFileChecksum(target);
  } catch (IOException e) {
    LOG.error("Unable to retrieve checksum for " + source + " or " + target, e);
  }
  return (sourceChecksum == null || targetChecksum == null ||
          sourceChecksum.equals(targetChecksum));
}