org.apache.storm.starter

Class SkewedRollingTopWords

java.lang.Object

org.apache.storm.topology.ConfigurableTopology

org.apache.storm.starter.SkewedRollingTopWords

public class SkewedRollingTopWords
extends ConfigurableTopology

This topology does a continuous computation of the top N words that the topology has seen in terms of cardinality. The top N computation is done in a completely scalable way, and a similar approach could be used to compute things like trending topics or trending images on Twitter. It takes an approach that assumes that some works will be much more common then other words, and uses partialKeyGrouping to better balance the skewed load.

Field Summary

Fields inherited from class org.apache.storm.topology.ConfigurableTopology

conf

Method Summary

Modifier and Type	Method and Description
static void	main(String[] args)
protected int	run(String[] args) Submits (runs) the topology.

Methods inherited from class org.apache.storm.topology.ConfigurableTopology

getConf, loadConf, start, submit, submit

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception

run

protected int run(String[] args)

Submits (runs) the topology.

Usage: "SkewedRollingTopWords [topology-name] [-local]"

By default, the topology is run locally under the name "slidingWindowCounts".

Examples:

``` # Runs in remote/cluster mode, with topology name "production-topology" $ storm jar storm-starter-jar-with-dependencies.jar org.apache.storm.starter.SkewedRollingTopWords production-topology ```

Specified by:

run in class ConfigurableTopology

Parameters:

args - First positional argument (optional) is topology name, second positional argument (optional) defines whether to run the topology locally ("-local") or remotely, i.e. on a real cluster