[
https://issues.apache.org/jira/browse/SPARK-21595?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16113388#comment-16113388
]
Herman van Hovell commented on SPARK-21595:
-------------------------------------------
The old and the new code are not exactly the same. The old code path would
start using a disk spilling buffer when a window would become larger than 4096
rows. The key difference is that old code path would not start to spill at that
point, that would only happen when the Spark would get pressed for memory and
the memory manager starts to force spills. The current version is overly active
and starts spilling at a much earlier stage. We have seen similar problems with
customer workloads on our end.
We either need to set this to a more sensible default, or return this to the
old behavior.
> introduction of spark.sql.windowExec.buffer.spill.threshold in spark 2.2
> breaks existing workflow
> -------------------------------------------------------------------------------------------------
>
> Key: SPARK-21595
> URL: https://issues.apache.org/jira/browse/SPARK-21595
> Project: Spark
> Issue Type: Bug
> Components: Documentation, PySpark
> Affects Versions: 2.2.0
> Environment: pyspark on linux
> Reporter: Stephan Reiling
> Priority: Minor
> Labels: documentation, regression
>
> My pyspark code has the following statement:
> {code:java}
> # assign row key for tracking
> df = df.withColumn(
> 'association_idx',
> sqlf.row_number().over(
> Window.orderBy('uid1', 'uid2')
> )
> )
> {code}
> where df is a long, skinny (450M rows, 10 columns) dataframe. So this creates
> one large window for the whole dataframe to sort over.
> In spark 2.1 this works without problem, in spark 2.2 this fails either with
> out of memory exception or too many open files exception, depending on memory
> settings (which is what I tried first to fix this).
> Monitoring the blockmgr, I see that spark 2.1 creates 152 files, spark 2.2
> creates >110,000 files.
> In the log I see the following messages (110,000 of these):
> {noformat}
> 17/08/01 08:55:37 INFO UnsafeExternalSorter: Spilling data because number of
> spilledRecords crossed the threshold 4096
> 17/08/01 08:55:37 INFO UnsafeExternalSorter: Thread 156 spilling sort data of
> 64.1 MB to disk (0 time so far)
> 17/08/01 08:55:37 INFO UnsafeExternalSorter: Spilling data because number of
> spilledRecords crossed the threshold 4096
> 17/08/01 08:55:37 INFO UnsafeExternalSorter: Thread 156 spilling sort data of
> 64.1 MB to disk (1 time so far)
> {noformat}
> So I started hunting for clues in UnsafeExternalSorter, without luck. What I
> had missed was this one message:
> {noformat}
> 17/08/01 08:55:37 INFO ExternalAppendOnlyUnsafeRowArray: Reached spill
> threshold of 4096 rows, switching to
> org.apache.spark.util.collection.unsafe.sort.UnsafeExternalSorter
> {noformat}
> Which allowed me to track down the issue.
> By changing the configuration to include:
> {code:java}
> spark.sql.windowExec.buffer.spill.threshold 2097152
> {code}
> I got it to work again and with the same performance as spark 2.1.
> I have workflows where I use windowing functions that do not fail, but took a
> performance hit due to the excessive spilling when using the default of 4096.
> I think to make it easier to track down these issues this config variable
> should be included in the configuration documentation.
> Maybe 4096 is too small of a default value?
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
