Examples with IFileInputStream org.apache.hadoop.mapred.IFileInputStream used on opensource projects

Example 1 with IFileInputStream

use of org.apache.hadoop.mapred.IFileInputStream in project hadoop by apache.

the class InMemoryMapOutput method doShuffle.

@Override
protected void doShuffle(MapHost host, IFileInputStream iFin, long compressedLength, long decompressedLength, ShuffleClientMetrics metrics, Reporter reporter) throws IOException {
    InputStream input = iFin;
    // Are map-outputs compressed?
    if (codec != null) {
        decompressor.reset();
        input = codec.createInputStream(input, decompressor);
    }
    try {
        IOUtils.readFully(input, memory, 0, memory.length);
        metrics.inputBytes(memory.length);
        reporter.progress();
        LOG.info("Read " + memory.length + " bytes from map-output for " + getMapId());
        /**
       * We've gotten the amount of data we were expecting. Verify the
       * decompressor has nothing more to offer. This action also forces the
       * decompressor to read any trailing bytes that weren't critical
       * for decompression, which is necessary to keep the stream
       * in sync.
       */
        if (input.read() >= 0) {
            throw new IOException("Unexpected extra bytes from input stream for " + getMapId());
        }
    } finally {
        CodecPool.returnDecompressor(decompressor);
    }
}

Example 2 with IFileInputStream

use of org.apache.hadoop.mapred.IFileInputStream in project hadoop by apache.

the class TestFetcher method testCorruptedIFile.

@Test
public void testCorruptedIFile() throws Exception {
    final int fetcher = 7;
    Path onDiskMapOutputPath = new Path(name.getMethodName() + "/foo");
    Path shuffledToDisk = OnDiskMapOutput.getTempPath(onDiskMapOutputPath, fetcher);
    fs = FileSystem.getLocal(job).getRaw();
    IFileWrappedMapOutput<Text, Text> odmo = new OnDiskMapOutput<Text, Text>(map1ID, mm, 100L, job, fetcher, true, fs, onDiskMapOutputPath);
    String mapData = "MAPDATA12345678901234567890";
    ShuffleHeader header = new ShuffleHeader(map1ID.toString(), 14, 10, 1);
    ByteArrayOutputStream bout = new ByteArrayOutputStream();
    DataOutputStream dos = new DataOutputStream(bout);
    IFileOutputStream ios = new IFileOutputStream(dos);
    header.write(dos);
    int headerSize = dos.size();
    try {
        ios.write(mapData.getBytes());
    } finally {
        ios.close();
    }
    int dataSize = bout.size() - headerSize;
    // Ensure that the OnDiskMapOutput shuffler can successfully read the data.
    MapHost host = new MapHost("TestHost", "http://test/url");
    ByteArrayInputStream bin = new ByteArrayInputStream(bout.toByteArray());
    try {
        // Read past the shuffle header.
        bin.read(new byte[headerSize], 0, headerSize);
        odmo.shuffle(host, bin, dataSize, dataSize, metrics, Reporter.NULL);
    } finally {
        bin.close();
    }
    // Now corrupt the IFile data.
    byte[] corrupted = bout.toByteArray();
    corrupted[headerSize + (dataSize / 2)] = 0x0;
    try {
        bin = new ByteArrayInputStream(corrupted);
        // Read past the shuffle header.
        bin.read(new byte[headerSize], 0, headerSize);
        odmo.shuffle(host, bin, dataSize, dataSize, metrics, Reporter.NULL);
        fail("OnDiskMapOutput.shuffle didn't detect the corrupted map partition file");
    } catch (ChecksumException e) {
        LOG.info("The expected checksum exception was thrown.", e);
    } finally {
        bin.close();
    }
    // Ensure that the shuffled file can be read.
    IFileInputStream iFin = new IFileInputStream(fs.open(shuffledToDisk), dataSize, job);
    try {
        iFin.read(new byte[dataSize], 0, dataSize);
    } finally {
        iFin.close();
    }
}