GitHub user cloud-fan opened a pull request:
https://github.com/apache/spark/pull/10498
[SPARK-12539][SQL][WIP] support writing bucketed table
Done:
* add bucket info in write path
* support writing bucketed `HadoopFsRelation`
TODO:
* figure out a way to store metadata for `HadoopFsRelation` without hive
meta store.
* figure out how to insert bucketed data into empty hive table. (currently
we just use hash map to [keep all
writers](https://github.com/apache/spark/blob/master/sql/hive/src/main/scala/org/apache/spark/sql/hive/hiveWriterContainers.scala#L249),
which maybe expensive when we have bucketing)
* when we append bucketed data to existing `HadoopFsRelation`, we may have
multi-files for one bucket, is that safe?
* hive support having bucketing without partitioning, should we support it?
* more safe check(fail fast if the given bucket info is wrong)
* code refine(a better way to delivery partitioning and bucketing info
though write path)
* add test(how to test it before implement read path?)
* add doc
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/cloud-fan/spark bucket-write
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/10498.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 #10498
----
commit 8cb24942890153bad70e003110a28d6111f0c407
Author: Wenchen Fan <[email protected]>
Date: 2015-12-28T12:15:34Z
write bucketed table
----
---
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]
