[
https://issues.apache.org/jira/browse/DRILL-4577?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15241886#comment-15241886
]
ASF GitHub Bot commented on DRILL-4577:
---------------------------------------
Github user hsuanyi commented on the pull request:
https://github.com/apache/drill/pull/461#issuecomment-210144139
In the background, Chun, Rahul and Jinfeng have done some experiments:
1. Functional-wise: If Hive uses storage-based impersonation, when a user
does "Show Tables" on Hive (without going through DRILL), this user will see
the tables which might not be under his/her permission. So from the aspect, the
behavior of Drill matches Hive with respect to visibility of tables displayed.
2. Performance-wise:
For INFORMATION_SCHEMA query against a hive instance with 1000 tables
(impersonation was off),
we got 38 seconds when the patch is not applied (non bulk load), and 8
seconds is when the patch is applied (bulk load).
> Improve performance for query on INFORMATION_SCHEMA when HIVE is plugged in
> ---------------------------------------------------------------------------
>
> Key: DRILL-4577
> URL: https://issues.apache.org/jira/browse/DRILL-4577
> Project: Apache Drill
> Issue Type: Improvement
> Components: Storage - Hive
> Reporter: Sean Hsuan-Yi Chu
> Assignee: Sean Hsuan-Yi Chu
> Fix For: 1.7.0
>
>
> A query such as
> {code}
> select * from INFORMATION_SCHEMA.`TABLES`
> {code}
> is converted as calls to fetch all tables from storage plugins.
> When users have Hive, the calls to hive metadata storage would be:
> 1) get_table
> 2) get_partitions
> However, the information regarding partitions is not used in this type of
> queries. Beside, a more efficient way is to fetch tables is to use
> get_multi_table call.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
