[
https://issues.apache.org/jira/browse/HIVE-474?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12913865#action_12913865
]
John Sichi commented on HIVE-474:
---------------------------------
Some comments after a brief look:
* The patch is going to need to be rebased against trunk (I guess after
HIVE-537 is committed)?
* We should make sure that in the case of a single distinct agg, we leave the
plan as it is today, and only use the new plan generation when multiple
distincts are present. This may already be the case; I couldn't quite tell
from the example plans in the test cases (it would be nice to have some simpler
queries for that).
* Regarding moving expression evaluation to the reduce side: in general, this
is something that needs cost-based optimization, due to factors like (a) data
size before and after expression evaluation and (b) parallelization benefit of
spreading out the computation over lots of mappers (assuming many more mappers
than reducers).
> Support for distinct selection on two or more columns
> -----------------------------------------------------
>
> Key: HIVE-474
> URL: https://issues.apache.org/jira/browse/HIVE-474
> Project: Hadoop Hive
> Issue Type: Improvement
> Components: Query Processor
> Reporter: Alexis Rondeau
> Assignee: Mafish
> Attachments: hive-474.0.4.2rc.patch
>
>
> The ability to select distinct several, individual columns as by example:
> select count(distinct user), count(distinct session) from actions;
> Currently returns the following failure:
> FAILED: Error in semantic analysis: line 2:7 DISTINCT on Different Columns
> not Supported user
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
