public void configure(JobConf job) {
this.jobConf = job;
urlNormalizers = new URLNormalizers(job, URLNormalizers.SCOPE_INJECT);
interval = jobConf.getInt("db.fetch.interval.default", 2592000);
filters = new URLFilters(jobConf);
scfilters = new ScoringFilters(jobConf);
scoreInjected = jobConf.getFloat("db.score.injected", 1.0f);
curTime = job.getLong("injector.current.time", System.currentTimeMillis());
}
/**
* Calculate how many maps to run. Number of maps is bounded by a minimum of the cumulative size
* of the copy / (distcp.bytes.per.map, default BYTES_PER_MAP or -m on the command line) and at
* most (distcp.max.map.tasks, default MAX_MAPS_PER_NODE * nodes in the cluster).
*
* @param totalBytes Count of total bytes for job
* @param job The job to configure
* @return Count of maps to run.
*/
private static void setMapCount(long totalBytes, JobConf job) throws IOException {
int numMaps = (int) (totalBytes / job.getLong(BYTES_PER_MAP_LABEL, BYTES_PER_MAP));
numMaps =
Math.min(
numMaps,
job.getInt(
MAX_MAPS_LABEL,
MAX_MAPS_PER_NODE * new JobClient(job).getClusterStatus().getTaskTrackers()));
job.setNumMapTasks(Math.max(numMaps, 1));
}
/**
* Produce splits such that each is no greater than the quotient of the total size and the
* number of splits requested.
*
* @param job The handle to the JobConf object
* @param numSplits Number of splits requested
*/
public InputSplit[] getSplits(JobConf job, int numSplits) throws IOException {
int cnfiles = job.getInt(SRC_COUNT_LABEL, -1);
long cbsize = job.getLong(TOTAL_SIZE_LABEL, -1);
String srcfilelist = job.get(SRC_LIST_LABEL, "");
if (cnfiles < 0 || cbsize < 0 || "".equals(srcfilelist)) {
throw new RuntimeException(
"Invalid metadata: #files("
+ cnfiles
+ ") total_size("
+ cbsize
+ ") listuri("
+ srcfilelist
+ ")");
}
Path src = new Path(srcfilelist);
FileSystem fs = src.getFileSystem(job);
FileStatus srcst = fs.getFileStatus(src);
ArrayList<FileSplit> splits = new ArrayList<FileSplit>(numSplits);
LongWritable key = new LongWritable();
FilePair value = new FilePair();
final long targetsize = cbsize / numSplits;
long pos = 0L;
long last = 0L;
long acc = 0L;
long cbrem = srcst.getLen();
SequenceFile.Reader sl = null;
try {
sl = new SequenceFile.Reader(fs, src, job);
for (; sl.next(key, value); last = sl.getPosition()) {
// if adding this split would put this split past the target size,
// cut the last split and put this next file in the next split.
if (acc + key.get() > targetsize && acc != 0) {
long splitsize = last - pos;
splits.add(new FileSplit(src, pos, splitsize, (String[]) null));
cbrem -= splitsize;
pos = last;
acc = 0L;
}
acc += key.get();
}
} finally {
checkAndClose(sl);
}
if (cbrem != 0) {
splits.add(new FileSplit(src, pos, cbrem, (String[]) null));
}
return splits.toArray(new FileSplit[splits.size()]);
}