[
https://issues.apache.org/jira/browse/DRILL-4455?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15256561#comment-15256561
]
ASF GitHub Bot commented on DRILL-4455:
---------------------------------------
Github user parthchandra commented on the pull request:
https://github.com/apache/drill/pull/398#issuecomment-214433058
I have two problems with this patch (apart from it being just too big to
review)
1) Building against 0.1-SNAPSHOT of a project that is in rapid development
is a courageous decision (I would rather not board a train while it is moving
at full speed). The current patch also does not build as there is no artifact
for Arrow 0.1-SNAPSHOT.
2) I see that the entire memory module of Drill is gone. I don't think that
drill memory management and accounting should be outsourced to Arrow. I can see
needing to use ArrowBuf for value vectors but removing the entire memory module
is not OK. I can see other uses of Direct/Heap memory (caching metadata for
example) that should be managed under a Drill resource manager. This ties back
to #1. Arrow needs to get its design figured out to allow Drill and other
projects to manage their resources while providing a 'reference'
implementation.
-1 until we resolve this.
> Depend on Apache Arrow for Vector and Memory
> --------------------------------------------
>
> Key: DRILL-4455
> URL: https://issues.apache.org/jira/browse/DRILL-4455
> Project: Apache Drill
> Issue Type: Bug
> Reporter: Steven Phillips
> Assignee: Steven Phillips
> Fix For: 2.0.0
>
>
> The code for value vectors and memory has been split and contributed to the
> apache arrow repository. In order to help this project advance, Drill should
> depend on the arrow project instead of internal value vector code.
> This change will require recompiling any external code, such as UDFs and
> StoragePlugins. The changes will mainly just involve renaming the classes to
> the org.apache.arrow namespace.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
