Processing Streaming Data - Real-Time Analytics - page 162

A Logging Bolt
The logging bolt is a very simple Bolt implementation that is useful
when testing topologies. All it does is print the results of an input tuple
to the console. This can be very useful when testing topologies, and it is
easy to implement with a BasicBolt . This bolt doesn't produce any
data, and just uses the fact that Tuple implements a reasonable
toString method:
public class LoggerBolt extends BaseBasicBolt {
private static final Logger LOG
= Logger. getLogger (LoggerBolt. class );
private static final long serialVersionUID = 1L;
public void execute(Tuple input,
BasicOutputCollector collector) {
LOG .info(input.toString());
}
public void
declareOutputFields(OutputFieldsDeclarer declarer) {
}
}
Implementing and Using Spouts
The Spout is a special form of Storm topology element that is responsible
for retrieving data from the outside world. Unlike Bolts , there is only an
IRichSpout interface to implement and no base class to derive from.
When a Spout is created, the open method is called first. This allows the
Spout to configure any connections to the outside world, such as
connectionstoqueueordatamotionservers.Likea Bolt ,the Spout should
capture its SpoutOutputCollector from the open method for later use:
public class EmptySpout implements IRichSpout {
transient SpoutOutputCollector collector;
public void open(Map conf, TopologyContext context,