[ 
https://issues.apache.org/jira/browse/BEAM-8794?focusedWorklogId=367085&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-367085
 ]

ASF GitHub Bot logged work on BEAM-8794:
----------------------------------------

                Author: ASF GitHub Bot
            Created on: 06/Jan/20 23:15
            Start Date: 06/Jan/20 23:15
    Worklog Time Spent: 10m 
      Work Description: apilloud commented on issue #10180: [BEAM-8794] 
Conditional aggregate project merge
URL: https://github.com/apache/beam/pull/10180#issuecomment-571357595
 
 
   run precommit
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


Issue Time Tracking
-------------------

    Worklog Id:     (was: 367085)
    Time Spent: 2h 10m  (was: 2h)

> Projects should be handled by an IOPushDownRule before applying 
> AggregateProjectMergeRule
> -----------------------------------------------------------------------------------------
>
>                 Key: BEAM-8794
>                 URL: https://issues.apache.org/jira/browse/BEAM-8794
>             Project: Beam
>          Issue Type: Improvement
>          Components: dsl-sql
>            Reporter: Kirill Kozlov
>            Assignee: Kirill Kozlov
>            Priority: Major
>          Time Spent: 2h 10m
>  Remaining Estimate: 0h
>
> It is more efficient to push-down projected fields at an IO level (vs merging 
> with an Aggregate), when supported.
> When running queries like:
> {code:java}
> select SUM(score) as total_score from <TABLE> group by name{code}
> Projects get merged with an aggregate, as a result Calc (after an 
> IOSourceRel) projects all fields and BeamIOPushDown rule does know what 
> fields can be dropped, thus not dropping any.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to