gistfile1.txt

import org.apache.hadoop.fs.*;
import org.apache.hadoop.conf.Configuration;
import java.util.Random;
import java.io.IOException;

public class RandomIO {

  private static class Reader extends Thread {
    Path path;
    FileSystem fs;
    volatile boolean done = false;
    int count = 0;

    public Reader(FileSystem fs, Path p) {
      this.fs = fs;
      this.path = p;
    }

    public void run() {
      try {
        internalRun();
      } catch (Exception e) {
        throw new RuntimeException(e);
      }
    }

    private void internalRun() throws Exception {
      long st = System.currentTimeMillis();
      long len = fs.getLength(path);
      FSDataInputStream dis = fs.open(path);
      Random r = new Random();
      System.out.println("Took " + (System.currentTimeMillis() - st) + " ms to open");

      byte buf[] = new byte[64*1024];
      while (!done) {
        long pos = r.nextInt((int)(len - buf.length));
        dis.readFully(pos, buf, 0, buf.length);
        count++;
      }

      System.out.println("Read " + count);

      dis.close();
    }
  }

  public static void main(String args[]) throws Exception {
    String file = args[0];
    Configuration conf = new Configuration();
    conf.setInt("io.file.buffer.size", 64*1024);
    conf.setInt("dfs.client.socketcache.capacity", 20);
    Path path = new Path(file);
    FileSystem fs = path.getFileSystem(conf);

    Reader r[] = new Reader[16];
    for (int i = 0; i < r.length; i++) {
      r[i] = new Reader(fs, path);
      r[i].start();
    }

    Thread.sleep(20000);
    for (Reader rdr : r) {
      rdr.done = true;
    }
    for (Reader rdr : r) {
      rdr.join();
    }
  }
}