java使用Storm Trident向cassandra插入行

Question

我试图在Cassandra 2.0.5，Storm版本0.9.0.1中向表插入一个简单的行

我的测试如下：

我有一个由id（int）和句子（text）列组成的表。id是主键

我的喷口生成句子，我添加一个ID（代码中的静态增量）

这是我的拓扑结构：

TridentTopology topology = new TridentTopology();
StateFactory cassandraStateFactory = CassandraMapState.nonTransactional(options);
Fields fields = new Fields("id", "sentence");
MyTridentTupleMapper tupleMapper = new MyTridentTupleMapper(keyspace, fields);
CassandraUpdater updater = new CassandraUpdater(tupleMapper);
TridentState wordCounts = topology.newStream("spout1", spout)
            .each(new Fields("sentence"), new AddId(), new Fields("id"))
            .partitionPersist(cassandraStateFactory, fields, updater);

LocalCluster cluster = new LocalCluster();
cluster.submitTopology("test", config, topology.build());    

MyTridentTupleMapper的代码：

https://github.com/guywald/trident-cassandra-read-write-examples/blob/master/src/test/java/com/guywald/storm/trident/cassandra/MyTridentTupleMapper.java

我得到以下异常：

2014-02-08 22:20:14 ERROR executor:0 - 
java.lang.RuntimeException: java.lang.ClassCastException: storm.trident.state.map.SnapshottableMap cannot be cast to com.hmsonline.storm.cassandra.trident.CassandraState
    at backtype.storm.utils.DisruptorQueue.consumeBatchToCursor(DisruptorQueue.java:90)
    at backtype.storm.utils.DisruptorQueue.consumeBatchWhenAvailable(DisruptorQueue.java:61)
    at backtype.storm.disruptor$consume_batch_when_available.invoke(disruptor.clj:62)
    at backtype.storm.daemon.executor$fn__3498$fn__3510$fn__3557.invoke(executor.clj:730)
    at backtype.storm.util$async_loop$fn__444.invoke(util.clj:403)
    at clojure.lang.AFn.run(AFn.java:24)
    at java.lang.Thread.run(Thread.java:744)

我不知道为什么它会返回此文件，并将感谢您的帮助

Answer 1

看起来CassandraUpdater期望CassandraState，而Cassandrampstate期望Cassandrampstate。非事务性创建不兼容的SnapshottableMap。 我相信常见的（或异国风情的）MapState更新程序将与CassandrampState一起使用。关于何时使用State vs MapState，这里有一个很好的解释：https://groups.google.com/forum/#!topic/storm-user/TASr2zWyzKs

我认为应该把CassandraStateFactory作为一个州工厂，让你的例子发挥作用