You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Jim Ancona (JIRA)" <ji...@apache.org> on 2011/07/09 15:35:16 UTC

[jira] [Commented] (CASSANDRA-2717) duplicate rows returned from SELECT where KEY term is duplicated

    [ https://issues.apache.org/jira/browse/CASSANDRA-2717?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13062376#comment-13062376 ] 

Jim Ancona commented on CASSANDRA-2717:
---------------------------------------

I added some tests to test/system/test_cql.py. I found that not only did "WHERE KEY = 'bar' and KEY = 'bar'" return two rows, so did "WHERE KEY = 'bar' and KEY = 'baz'" and "WHERE KEY IN ('bar', 'bar')"

The attached patch makes having more than one "KEY =" clause be an error, and changes the List of keys in WhereClause to a Set. 

Jonathan mentioned that OR support was added, but I didn't see that in cassandra-0.8. Am I looking at the wrong branch? If so, this patch will have to be reworked, along with the logic in WhereClause.

> duplicate rows returned from SELECT where KEY term is duplicated
> ----------------------------------------------------------------
>
>                 Key: CASSANDRA-2717
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2717
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.8.0 beta 2
>            Reporter: Aaron Morton
>            Assignee: Jim Ancona
>            Priority: Minor
>              Labels: cql, lhf
>         Attachments: v1-0001-CASSANDRA-2717-Prevent-multiple-KEY-terms-properly-han.txt
>
>
> Noticed while working on CASSANDRA-2268 when random keys generated during a mutli_get test contain duplicate keys. 
> The thrift multiget_slice() returns only the unique rows because of the map generated for the result. 
> CQL will return a row for each KEY term in the SELECT. 
> I could make QueryProcessor.getSlice() only create commands for the unique keys if we wanted to. 
> Not sure it's a bug and it's definitely not something that should come up to often, reporting it because it's different to the thrift mutli_get operation. 
> Happy to close if it's by design. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira