[
https://issues.apache.org/jira/browse/HIVE-26072?focusedWorklogId=794629&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-794629
]
ASF GitHub Bot logged work on HIVE-26072:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 24/Jul/22 14:11
Start Date: 24/Jul/22 14:11
Worklog Time Spent: 10m
Work Description: ayushtkn opened a new pull request, #3142:
URL: https://github.com/apache/hive/pull/3142
After https://issues.apache.org/jira/browse/HIVE-23530 , almost all compute
stats functions are vectorizable. Only function that is not vectorizable is
"compute_bit_vector" for ndv statistics computation. This causes "create table
as select" and "insert overwrite select" queries to run in non-vectorized mode.
Even a very naive implementation of vectorized compute_bit_vector gives
about 50% performance improvement on simple "insert overwrite select" queries.
That is because entire mapper or reducer can run in vectorized mode.
Issue Time Tracking
-------------------
Worklog Id: (was: 794629)
Time Spent: 1h 20m (was: 1h 10m)
> Enable vectorization for stats gathering (tablescan op)
> -------------------------------------------------------
>
> Key: HIVE-26072
> URL: https://issues.apache.org/jira/browse/HIVE-26072
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Reporter: Rajesh Balamohan
> Assignee: Ayush Saxena
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> https://issues.apache.org/jira/browse/HIVE-24510 enabled vectorization for
> compute_bit_vector.
> But tablescan operator for stats gathering is disabled by default.
> [https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/optimizer/physical/Vectorizer.java#L2577]
> Need to enable vectorization for this. This can significantly reduce runtimes
> for analyze statements for large tables.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
