[
https://issues.apache.org/jira/browse/SPARK-12837?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15963510#comment-15963510
]
Ruslan Dautkhanov edited comment on SPARK-12837 at 4/10/17 9:29 PM:
--------------------------------------------------------------------
It might be a bug in broadcast join.
Following Spark 2 query fails with
{quote}Total size of serialized results of 128 tasks (1026.2 MB) is bigger than
spark.driver.maxResultSize (1024.0 MB){quote}
when we set
{code}
sqlc.setConf("spark.sql.autoBroadcastJoinThreshold", 500 *1024*1024) # 500 mb
{code}
{noformat}
SELECT . . . <skip>
m.year,
m.quarter,
t.individ,
t.hh_key
FROM mv_dev.mv_raw_all_20170314 m,
disc_dv.tsp_dv_02122017 t
where m.psn = t.person_seq_no
limit 10
{noformat}
when we drop down to 400Mb
{code}
sqlc.setConf("spark.sql.autoBroadcastJoinThreshold", 400 *1024*1024) # 400 mb
{code}
, this error does not show up.
was (Author: tagar):
It might be a bug in broadcast join.
Following Spark 2 query fails with
Total size of serialized results of 128 tasks (1026.2 MB) is bigger than
spark.driver.maxResultSize (1024.0 MB)
when we set
{code}
sqlc.setConf("spark.sql.autoBroadcastJoinThreshold", 500 *1024*1024) # 500 mb
{code}
{noformat}
SELECT . . . <skip>
m.year,
m.quarter,
t.individ,
t.hh_key
FROM mv_dev.mv_raw_all_20170314 m,
disc_dv.tsp_dv_02122017 t
where m.psn = t.person_seq_no
limit 10
{noformat}
when we drop down to 400Mb
{code}
sqlc.setConf("spark.sql.autoBroadcastJoinThreshold", 400 *1024*1024) # 400 mb
{code}
, this error does not show up.
> Spark driver requires large memory space for serialized results even there
> are no data collected to the driver
> --------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-12837
> URL: https://issues.apache.org/jira/browse/SPARK-12837
> Project: Spark
> Issue Type: Question
> Components: SQL
> Affects Versions: 1.5.2, 1.6.0
> Reporter: Tien-Dung LE
> Assignee: Wenchen Fan
> Priority: Critical
> Fix For: 2.0.0
>
>
> Executing a sql statement with a large number of partitions requires a high
> memory space for the driver even there are no requests to collect data back
> to the driver.
> Here are steps to re-produce the issue.
> 1. Start spark shell with a spark.driver.maxResultSize setting
> {code:java}
> bin/spark-shell --driver-memory=1g --conf spark.driver.maxResultSize=1m
> {code}
> 2. Execute the code
> {code:java}
> case class Toto( a: Int, b: Int)
> val df = sc.parallelize( 1 to 1e6.toInt).map( i => Toto( i, i)).toDF
> sqlContext.setConf( "spark.sql.shuffle.partitions", "200" )
> df.groupBy("a").count().saveAsParquetFile( "toto1" ) // OK
> sqlContext.setConf( "spark.sql.shuffle.partitions", 1e3.toInt.toString )
> df.repartition(1e3.toInt).groupBy("a").count().repartition(1e3.toInt).saveAsParquetFile(
> "toto2" ) // ERROR
> {code}
> The error message is
> {code:java}
> Caused by: org.apache.spark.SparkException: Job aborted due to stage failure:
> Total size of serialized results of 393 tasks (1025.9 KB) is bigger than
> spark.driver.maxResultSize (1024.0 KB)
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
