[
https://issues.apache.org/jira/browse/CASSANDRA-6963?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jonathan Ellis resolved CASSANDRA-6963.
---------------------------------------
Resolution: Duplicate
Fix Version/s: (was: 2.0.6)
should be fixed in current 2.0 HEAD as of CASSANDRA-6793
> hadoop_cql3_wordcount example : inconsistent schema definitions means writing
> results to Cassandra fails.
> ---------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-6963
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6963
> Project: Cassandra
> Issue Type: Bug
> Components: Examples
> Environment: Ubuntu 12.04, Java version 1.7.0_51
> Reporter: yizhou
> Priority: Minor
> Attachments: Screen Shot 2014-03-31 at 5.28.26 PM.png
>
> Original Estimate: 1h
> Remaining Estimate: 1h
>
> The schema defined in WordCountSetup.java specifies 'row_id' as primary key
> for output_words. Ultimately, this is inconsistent with the assumed schema in
> WordCount.java. What ends up happening is that within the
> O.A.C.hadoop.cql3.CqlRecordWriter class, the getPartitionKey function returns
> null when trying looking up the primary key as 'row_id'. This propagates
> eventually causing a NullPointerException. The results of wordcount are not
> recorded in Cassandra as a result.
> This error can be reproduced on cassandra 2.0.6 by building cassandra,
> building the hadoop_cql3_wordcount example, and then running word_count_setup
> and word_count.
> I attached a patch which modifies the output_words schema so that every runs
> out of the box.
> I attached a screenshot of the exception stack.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
