GitHub user rajeshbalamohan opened a pull request:
https://github.com/apache/spark/pull/10861
SPARK-12948. [SQL]. Consider reducing size of broadcasts in OrcRelation
Size of broadcasted data in OrcRelation was significantly higher when
running query with large number of partitions (e.g TPC-DS). And it has an
impact on the job runtime. This would be more evident when there is large
number of partitions/splits. Profiler snapshot is attached in SPARK-12948
(https://issues.apache.org/jira/secure/attachment/12783513/SPARK-12948_cpuProf.png).
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/rajeshbalamohan/spark SPARK-12948
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/10861.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 #10861
----
commit 4da7a22d2195c77e27aa4f3aa957b1fdc0d57f5b
Author: Rajesh Balamohan <[email protected]>
Date: 2016-01-21T05:09:09Z
SPARK-12948. [SQL]. Consider reducing size of broadcasts in OrcRelation
----
---
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]
