[
https://issues.apache.org/jira/browse/SPARK-24906?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17006691#comment-17006691
]
Lior Chaga commented on SPARK-24906:
------------------------------------
Looking at the PR, I see only using configurable estimations for Struct, Map
and Array types, not something that is based on row group or any sample method
of real data.
Is there an additional development effort on this ticket?
I'd love to try and take part with my limited scala skills.
Anyway, not every heuristic is suitable for every use case. Rough estimations
like in the attached PR are not good for complex schemas or even data sources
that have many null values in participating columns.
On the other hand, using sample of metadata has a higher cost, and might be an
overkill for the naive cases.
Perhaps solution might be introducing a pluggable estimation strategy,
providing the naive implementation but allowing spark users to provide their
own strategy?
> Adaptively set split size for columnar file to ensure the task read data size
> fit expectation
> ---------------------------------------------------------------------------------------------
>
> Key: SPARK-24906
> URL: https://issues.apache.org/jira/browse/SPARK-24906
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.3.1
> Reporter: Jason Guo
> Priority: Major
> Attachments: image-2018-07-24-20-26-32-441.png,
> image-2018-07-24-20-28-06-269.png, image-2018-07-24-20-29-24-797.png,
> image-2018-07-24-20-30-24-552.png
>
>
> For columnar file, such as, when spark sql read the table, each split will be
> 128 MB by default since spark.sql.files.maxPartitionBytes is default to
> 128MB. Even when user set it to a large value, such as 512MB, the task may
> read only few MB or even hundreds of KB. Because the table (Parquet) may
> consists of dozens of columns while the SQL only need few columns. And spark
> will prune the unnecessary columns.
>
> In this case, spark DataSourceScanExec can enlarge maxPartitionBytes
> adaptively.
> For example, there is 40 columns , 20 are integer while another 20 are long.
> When use query on an integer type column and an long type column, the
> maxPartitionBytes should be 20 times larger. (20*4+20*8) / (4+8) = 20.
>
> With this optimization, the number of task will be smaller and the job will
> run faster. More importantly, for a very large cluster (more the 10 thousand
> nodes), it will relieve RM's schedule pressure.
>
> Here is the test
>
> The table named test2 has more than 40 columns and there are more than 5 TB
> data each hour.
> When we issue a very simple query
>
> {code:java}
> select count(device_id) from test2 where date=20180708 and hour='23'{code}
>
> There are 72176 tasks and the duration of the job is 4.8 minutes
> !image-2018-07-24-20-26-32-441.png!
>
> Most tasks last less than 1 second and read less than 1.5 MB data
> !image-2018-07-24-20-28-06-269.png!
>
> After the optimization, there are only 1615 tasks and the job last only 30
> seconds. It almost 10 times faster.
> !image-2018-07-24-20-29-24-797.png!
>
> The median of read data is 44.2MB.
> !image-2018-07-24-20-30-24-552.png!
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
