GitHub user chenghao-intel opened a pull request:
https://github.com/apache/spark/pull/2589
[SPARK-3739] [SQL] Update the split num base on block size for table
scanning
Source file input split is probably better based on block size of HDFS,
while scanning Hive table, other than the settings of 'mapred.reducer.tasks' or
'taskScheduler.defaultParallelism'. Checkout
http://grepcode.com/file/repository.cloudera.com/content/repositories/releases/org.apache.hadoop/hadoop-core/2.3.0-mr1-cdh5.1.0/org/apache/hadoop/mapred/FileInputFormat.java#203
for how Hadoop handles the input splits.
Currently, too many splits for small source table file, which may causes
performance issue, as well as the different result of Hive in unit test.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/chenghao-intel/spark source_split
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/2589.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #2589
----
commit 5f0d75b8cdebd6c63b9a6623055742cd0e019aa7
Author: Cheng Hao <[email protected]>
Date: 2014-09-30T03:10:52Z
Update the split num base on block size for table scanning
----
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
