[
https://issues.apache.org/jira/browse/CASSANDRA-745?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12806145#action_12806145
]
Stu Hood commented on CASSANDRA-745:
------------------------------------
This patch sets finishKey to the minimum token if the start_key is equal to the
empty string, which probably indicates that it needs a lot more testing.
Also, I don't think this will be very useful without some API changes. Yes,
someone using RandomPartitioner will be able to pass start="" and finish="" to
get all of the rows in their cluster, but that would be an extremely
inefficient way to get all rows. Additionally, if they use anything other than
the empty string, the behaviour becomes undefined: MD5s won't sort the same way
that strings do, so passing in strings is a guessing game.
If this is supposed to be for Hadoop integration, I suggest that we try to
build an efficient API call specifically for that task, including locality
information, and dividing the keyspace into a certain number of input
partitions.
> Support Range Queries all the partitioners currently supported.
> ---------------------------------------------------------------
>
> Key: CASSANDRA-745
> URL: https://issues.apache.org/jira/browse/CASSANDRA-745
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Affects Versions: 0.6
> Environment: Linux, Cassandra trunk
> Reporter: Vijay
> Assignee: Vijay
> Fix For: 0.6
>
> Attachments: 745-Patch.txt
>
>
> Due to the changes which was made to incorprate Decorated keys in cassandra
> we can add this support with some minor changes. Basically this patch will
> support the Range queries on the keyspaces.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
