[
https://issues.apache.org/jira/browse/CASSANDRA-3064?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13087940#comment-13087940
]
Jonathan Ellis commented on CASSANDRA-3064:
-------------------------------------------
The difference introduced by 2498 is, if it knows "I'm looking for exactly
{C0,C1,C2,C3,C4,C5}, then once it finds those columns in a sstable, such that
the maximum (client) timestamp in the other sstables is less than the values it
found, it doesn't have to look at those others.
But if you give it a range it can't assume that there is no C00, for instance,
in an older sstable that it might need to include.
> Add query-by-column mode to stress.java
> ---------------------------------------
>
> Key: CASSANDRA-3064
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3064
> Project: Cassandra
> Issue Type: Task
> Components: Tools
> Reporter: Jonathan Ellis
> Assignee: Pavel Yaskevich
> Priority: Minor
> Fix For: 1.0
>
>
> This is necessary to be able to demonstrate CASSANDRA-2498 as well as any
> other differences in slice-based vs name-based queries.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
