[
https://issues.apache.org/jira/browse/HIVE-15112?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15704442#comment-15704442
]
Ferdinand Xu commented on HIVE-15112:
-------------------------------------
[~csun], thank you for your review. I tried to debug the vectorization engine
and found it doesn't support struct type at this point while HIVE-11212 only
created the types instead of supporting it. Detail is available at
[L1957|https://github.com/apache/hive/blob/6536e30f2ee9707e16d4cd8d7c8321c4073a30b9/ql/src/java/org/apache/hadoop/hive/ql/optimizer/physical/Vectorizer.java#L1957]
and [L195 |
https://github.com/apache/hive/blob/6536e30f2ee9707e16d4cd8d7c8321c4073a30b9/ql/src/java/org/apache/hadoop/hive/ql/optimizer/physical/Vectorizer.java#L195]
from vectorizer class. Now HIVE-15112 is only focusing on Parquet vectorized
reader and UT has been passed. I am not going to add ptest in this PR since for
struct type vectorization execution processes it row by row. We will file other
ticket to address that.
https://cwiki.apache.org/confluence/display/Hive/Vectorized+Query+Execution
> Implement Parquet vectorization reader for Struct type
> ------------------------------------------------------
>
> Key: HIVE-15112
> URL: https://issues.apache.org/jira/browse/HIVE-15112
> Project: Hive
> Issue Type: Sub-task
> Reporter: Ferdinand Xu
> Assignee: Ferdinand Xu
>
> Like HIVE-14815, we need support Parquet vectorized reader for struct type.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
