[
https://issues.apache.org/jira/browse/SPARK-14008?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15628400#comment-15628400
]
Hyukjin Kwon commented on SPARK-14008:
--------------------------------------
Hi [~sameerag], I just wonder if this might be supposed to be resolved as Done
(I assume this is a parent task as a container and it seems sub-tasks are
done). I just read this
{quote}
* Umbrellas are frequently marked Done if they are just container issues that
don't correspond to an actionable change of their own
{quote}
from wiki :).
> Cleanup/Extend the Vectorized Parquet Reader
> --------------------------------------------
>
> Key: SPARK-14008
> URL: https://issues.apache.org/jira/browse/SPARK-14008
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Reporter: Sameer Agarwal
>
> Per [~nongli]'s suggestions:
> We should do these things:
> 1. Remove the non-vectorized parquet reader code.
> 2. Support the remaining types (just big decimals?)
> 3. Move the logic to determine if our parquet reader can be used to planning.
> Only complex types should fall back to the parquet-mr reader.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
