[
https://issues.apache.org/jira/browse/HIVE-8869?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14215100#comment-14215100
]
Gopal V commented on HIVE-8869:
-------------------------------
bq. It seems some readers (Accumulo, Vectorization) assume that the schema
should contain all the columns of the tuples that are read, even if the columns
are pruned
Vectorization only assumes that the column offsets remain identical. They still
require pruning of columns when reading data through.
Does this break the column projection into the ORC layer (it gets a bit-set of
true/false to read or not)?
> RowSchema not updated for some ops when columns are pruned
> ----------------------------------------------------------
>
> Key: HIVE-8869
> URL: https://issues.apache.org/jira/browse/HIVE-8869
> Project: Hive
> Issue Type: Bug
> Affects Versions: 0.14.0, 0.15.0
> Reporter: Jesús Camacho Rodríguez
> Assignee: Jesús Camacho Rodríguez
> Attachments: HIVE-8869.01.patch, HIVE-8869.02.patch, HIVE-8869.patch
>
>
> When columns are pruned in ColumnPrunerProcFactory, updating the row schema
> behavior is not consistent among operators: some will update their RowSchema,
> while some others will not.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
