Basic Storm topology example that updates DB persistent state with correct error handling. The code is based on Storm 0.9.0.1, Cassandra 2.0.4 and Java 7.
The [first release] ( https://github.com/svendx4f/stormRoomOccupancy/releases/tag/v1.0.1) is explained in great details in my blog post on [scalable real-time state update with Storm] (http://svendvanderveken.wordpress.com/2013/07/30/scalable-real-time-state-update-with-storm/)
The current code is an update explained in my blog post on [Storm error handling] ( http://svendvanderveken.wordpress.com/2014/02/01/notes-on-storm-trident-error-handling)
In order to run this example, an instance of Cassandra with the following key space is required:
CREATE KEYSPACE EVENT_POC WITH replication = { 'class': 'SimpleStrategy', 'replication_factor': '1' } ;
(tables are re-created everytime time the topology is re-deployed)
Maybe edit this line in Deployer.java if you Nimbus is not reachable on that IP:
config.put("nimbus.host" , "192.168.33.10");
Then package the topology:
mvn package
And deploy it:
storm jar target/stormRoomOccupancy-0.0.2-SNAPSHOT-jar-with-dependencies.jar svend.storm.example.conference.Deployer