Example usage for org.apache.hadoop.mapred JobConf getJobLocalDir

Introduction

In this page you can find the example usage for org.apache.hadoop.mapred JobConf getJobLocalDir.

Prototype

public String getJobLocalDir() 

Source Link

Document

Get job-specific shared directory for use as scratch space

When a job starts, a shared directory is created at location ${mapreduce.cluster.local.dir}/taskTracker/$user/jobcache/$jobid/work/ .

Usage

From source file:org.broadinstitute.sting.gatk.hadoop.LociRecordReader.java

License:Open Source License

public void initialize(InputSplit spl, TaskAttemptContext ctx) throws IOException {
    final FileVirtualSplit split = (FileVirtualSplit) spl;

    file = split.getPath();// ww  w .j  av a2  s  . co m
    fs = file.getFileSystem(ctx.getConfiguration());

    final FSDataInputStream in = fs.open(file);
    codec = new BAMRecordCodec(new SAMFileReader(in).getFileHeader());

    in.seek(0);
    bci = new BlockCompressedInputStream(
            new WrapSeekable<FSDataInputStream>(in, fs.getFileStatus(file).getLen(), file));

    virtualStart = split.getStartVirtualOffset();
    fileStart = virtualStart >>> 16;
    virtualEnd = split.getEndVirtualOffset();
    fileEnd = virtualEnd >>> 16;
    idx = new SplittingBAMIndex(file.getFileSystem(ctx.getConfiguration()).open(getIdxPath(file)));
    codec.setInputStream(bci);
    bci.seek(virtualStart);

    JobConf job = new JobConf(ctx.getConfiguration());
    jobDir = new String(job.getJobLocalDir());
    attemptID = ctx.getTaskAttemptID().toString();
}

---