Mailing-List: contact issues-help@spark.apache.org; run by ezmlm
Precedence: bulk
Date: Thu, 21 Sep 2017 05:39:00 +0000 (UTC)
From: "StanZhai (JIRA)" <jira@apache.org>
To: issues@spark.apache.org
Message-ID: <JIRA.13103817.1505972298000.163763.1505972340138@Atlassian.JIRA>
In-Reply-To: <JIRA.13103817.1505972298000@Atlassian.JIRA>
References: <JIRA.13103817.1505972298000@Atlassian.JIRA> <JIRA.13103817.1505972298505@jira-lw-us.apache.org>
Subject: [jira] [Updated] (SPARK-22084) Performance regression in
 aggregation strategy
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Thu, 21 Sep 2017 05:39:05 -0000


     [ https://issues.apache.org/jira/browse/SPARK-22084?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

StanZhai updated SPARK-22084:
-----------------------------
    Description: 
{code:sql}
SELECT a, SUM(b) AS b0, SUM(b) AS b1 
FROM VALUES(1, 1), (2, 2) AS (a, b) 
GROUP BY a
{code}

Two exactly the same SUM(b) expressions in the SQL, and the following is the physical plan in Spark 2.x.

{code}
== Physical Plan ==
*HashAggregate(keys=[a#11], functions=[sum(cast(b#12 as bigint)), sum(cast(b#12 as bigint))])
+- Exchange hashpartitioning(a#11, 200)
   +- *HashAggregate(keys=[a#11], functions=[partial_sum(cast(b#12 as bigint)), partial_sum(cast(b#12 as bigint))])
      +- LocalTableScan [a#11, b#12]
{code}

functions in Aggregate should be: functions=[partial_sum(cast(b#12 as bigint))]

  was:
{code:sql}
SELECT a, SUM(b) AS b0, SUM(b) AS b1 
FROM VALUES(1, 1), (2, 2) AS (a, b) 
GROUP BY a
{code}

Two exactly the same SUM(b) expressions in the SQL, and the following is the physical plan in Spark 2.x.

== Physical Plan ==
*HashAggregate(keys=[a#11], functions=[sum(cast(b#12 as bigint)), sum(cast(b#12 as bigint))])
+- Exchange hashpartitioning(a#11, 200)
   +- *HashAggregate(keys=[a#11], functions=[partial_sum(cast(b#12 as bigint)), partial_sum(cast(b#12 as bigint))])
      +- LocalTableScan [a#11, b#12]

functions in Aggregate should be: functions=[partial_sum(cast(b#12 as bigint))]


> Performance regression in aggregation strategy
> ----------------------------------------------
>
>                 Key: SPARK-22084
>                 URL: https://issues.apache.org/jira/browse/SPARK-22084
>             Project: Spark
>          Issue Type: Improvement
>          Components: SQL
>    Affects Versions: 2.0.0, 2.0.1, 2.0.2, 2.1.0, 2.1.1, 2.2.0
>            Reporter: StanZhai
>
> {code:sql}
> SELECT a, SUM(b) AS b0, SUM(b) AS b1 
> FROM VALUES(1, 1), (2, 2) AS (a, b) 
> GROUP BY a
> {code}
> Two exactly the same SUM(b) expressions in the SQL, and the following is the physical plan in Spark 2.x.
> {code}
> == Physical Plan ==
> *HashAggregate(keys=[a#11], functions=[sum(cast(b#12 as bigint)), sum(cast(b#12 as bigint))])
> +- Exchange hashpartitioning(a#11, 200)
>    +- *HashAggregate(keys=[a#11], functions=[partial_sum(cast(b#12 as bigint)), partial_sum(cast(b#12 as bigint))])
>       +- LocalTableScan [a#11, b#12]
> {code}
> functions in Aggregate should be: functions=[partial_sum(cast(b#12 as bigint))]


--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org