[
https://issues.apache.org/jira/browse/CASSANDRA-5805?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13719868#comment-13719868
]
Aleksey Yeschenko commented on CASSANDRA-5805:
----------------------------------------------
Maps and Lists are also affected, but I'm not certain that the issue is
necessarily collection-related.
> CQL 'set' returns incorrect value
> ---------------------------------
>
> Key: CASSANDRA-5805
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5805
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 2.0 beta 2
> Reporter: Yuki Morishita
> Priority: Critical
>
> CQL 'set' returns incorrect value after flush.
> Create the following table:
> {code}
> CREATE KEYSPACE ks WITH replication = {'class': 'SimpleStrategy',
> 'replication_factor': 1};
> USE ks;
> CREATE TABLE cf ( k int PRIMARY KEY , s set<int> );
> {code}
> Insert data:
> {code}
> INSERT INTO cf (k, s) VALUES (1, {1});
> INSERT INTO cf (k, s) VALUES (1, {2});
> {code}
> This should return:
> {code}
> cqlsh:ks> SELECT * FROM cf;
> k | s
> ---+--------
> 1 | {2}
> {code}
> and it does when no flush has happened.
> But when I do flush after each insert, it starts returning:
> {code}
> cqlsh:ks> SELECT * FROM cf;
> k | s
> ---+--------
> 1 | {1, 2}
> {code}
> 'system.local' table flushes every time it inserts(updates) tokens, and this
> behavior is causing 'nodetool move' to act weirdly.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
