[
https://issues.apache.org/jira/browse/DRILL-3940?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15001468#comment-15001468
]
Jacques Nadeau commented on DRILL-3940:
---------------------------------------
I spent some more time looking at your allocator branch. It seems like there
are a number of places where you've added AutoCloseable. My feedback above was
to the title of this JIRA which only focused on the RecordBatch interface.
I think the autocloseable additions you made to other classes, such as those to
SortRecordBatchBuilder and WritableBatch are very useful and whole-heartedly
support those additions.
On the RecordBatch situation specifically, I don't think we need to add a new
interface. All the implementations are AutoCloseable and this functionality is
exposed in the place where their state should be managed (BaseRootExec,
specifically [1]),. Exposing the close method to other operators breaks
operator lifecycle encapsulation.
[1]
https://github.com/apache/drill/blob/master/exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/BaseRootExec.java#L140
> Make RecordBatch AutoCloseable
> ------------------------------
>
> Key: DRILL-3940
> URL: https://issues.apache.org/jira/browse/DRILL-3940
> Project: Apache Drill
> Issue Type: Bug
> Components: Execution - Flow
> Affects Versions: 1.2.0
> Reporter: Chris Westin
> Assignee: Chris Westin
>
> This made it easier to find RecordBatches that were not cleaned up (because
> the compiler complains about AutoCloseable resources that haven't been
> closed).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
