[jira] [Commented] (HIVE-6604) Fix vectorized input to work with ACID

["Jitendra Nath Pandey (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-6604?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13960816#comment-13960816 ] 

Jitendra Nath Pandey commented on HIVE-6604:
--------------------------------------------

{code} 
+    if (!innerReader.next(key, value)) {
+      return false;
+    }
{code}
innerReader.hasNext() should be used, we might loose a row when 'next' is called again.

bq. VectorizedBatchUtil.AddRowToBatch
This method has been used only for RCFile, which is not yet supported in vectorization, and has not really been maintained e.g. It doesn't support Decimal, DATE types. However, I think it should be simple to update this method to handle Decimal and DATE as well.

bq. static VectorizedRowBatch createRowBatch(VectorizedRowBatchCtx ctx) 
Is this really needed, rowBatchCtx.createVectorizedRowBatch() can be called directly?

It would be great to have an end to end (.q) test with ACID and vectorization on, however that might be too complicated to do in this jira.


> Fix vectorized input to work with ACID
> --------------------------------------
>
>                 Key: HIVE-6604
>                 URL: https://issues.apache.org/jira/browse/HIVE-6604
>             Project: Hive
>          Issue Type: Sub-task
>            Reporter: Owen O'Malley
>            Assignee: Owen O'Malley
>            Priority: Blocker
>             Fix For: 0.13.0
>
>         Attachments: HIVE-6604.patch
>
>
> Fix the VectorizedOrcInputFormat to work with the ACID directories.



--
This message was sent by Atlassian JIRA
(v6.2#6252)