RealtimeIOConfig realtimeIOConfig = spec.getIOConfig();
RealtimeTuningConfig tuningConfig = spec.getTuningConfig()
.withBasePersistDirectory(new File(toolbox.getTaskWorkDir(), "persist"))
.withVersioningPolicy(versioningPolicy);
final FireDepartment fireDepartment = new FireDepartment(
dataSchema,
realtimeIOConfig,
tuningConfig,
null,
null,
null,
null
);
final RealtimeMetricsMonitor metricsMonitor = new RealtimeMetricsMonitor(ImmutableList.of(fireDepartment));
this.queryRunnerFactoryConglomerate = toolbox.getQueryRunnerFactoryConglomerate();
// NOTE: This pusher selects path based purely on global configuration and the DataSegment, which means
// NOTE: that redundant realtime tasks will upload to the same location. This can cause index.zip and
// NOTE: descriptor.json to mismatch, or it can cause historical nodes to load different instances of the
// NOTE: "same" segment.
final RealtimePlumberSchool plumberSchool = new RealtimePlumberSchool(
toolbox.getEmitter(),
toolbox.getQueryRunnerFactoryConglomerate(),
toolbox.getSegmentPusher(),
lockingSegmentAnnouncer,
segmentPublisher,
toolbox.getNewSegmentServerView(),
toolbox.getQueryExecutorService(),
null,
null,
null,
null,
null,
null,
0
);
this.plumber = plumberSchool.findPlumber(dataSchema, tuningConfig, fireDepartment.getMetrics());
try {
plumber.startJob();
// Set up metrics emission
toolbox.getMonitorScheduler().addMonitor(metricsMonitor);
// Time to read data!
long nextFlush = new DateTime().plus(intermediatePersistPeriod).getMillis();
while (firehose.hasMore()) {
final InputRow inputRow;
try {
inputRow = firehose.nextRow();
if (inputRow == null) {
continue;
}
int currCount = plumber.add(inputRow);
if (currCount == -1) {
fireDepartment.getMetrics().incrementThrownAway();
log.debug("Throwing away event[%s]", inputRow);
if (System.currentTimeMillis() > nextFlush) {
plumber.persist(firehose.commit());
nextFlush = new DateTime().plus(intermediatePersistPeriod).getMillis();
}
continue;
}
fireDepartment.getMetrics().incrementProcessed();
if (currCount >= tuningConfig.getMaxRowsInMemory() || System.currentTimeMillis() > nextFlush) {
plumber.persist(firehose.commit());
nextFlush = new DateTime().plus(intermediatePersistPeriod).getMillis();
}
}
catch (ParseException e) {
log.warn(e, "unparseable line");
fireDepartment.getMetrics().incrementUnparseable();
}
}
}
catch (Throwable e) {
normalExit = false;