[jira] [Commented] (DRILL-2719) Behavior of ValueVector#getBuffers(clear) is inconsistent across VV types

["Chris Westin (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DRILL-2719?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14486501#comment-14486501 ] 

Chris Westin commented on DRILL-2719:
-------------------------------------

So what's your solution going to be? I've run into this with super-strict accounting on the new allocator, and am doing some work in this area now myself -- I'd prefer not to collide.

> Behavior of ValueVector#getBuffers(clear) is inconsistent across VV types
> -------------------------------------------------------------------------
>
>                 Key: DRILL-2719
>                 URL: https://issues.apache.org/jira/browse/DRILL-2719
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Execution - Data Types
>    Affects Versions: 0.8.0
>            Reporter: Hanifi Gunes
>            Assignee: Hanifi Gunes
>             Fix For: 0.9.0
>
>
> The major cause of inconsistency here is that reference count does not change upon a call to BaseDataVV#getBuffers. However, it does change when the method is invoked on a composite vector type. See for instance VariableLengthVV[1] or a NullableVV[2].
> The proposal is to make this behavior consistent first and later we can consider separating out the logic that clears a vector from the logic that returns underlying buffers.
> [1]: https://github.com/apache/drill/blob/master/exec/java-exec/src/main/codegen/templates/VariableLengthVectors.java#L151
> [2]: https://github.com/apache/drill/blob/master/exec/java-exec/src/main/codegen/templates/NullableValueVectors.java#L82



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)