[jira] [Updated] (HIVE-16807) Persist row and column stats for HDFS directory across multiple external tables

["Mala Chikka Kempanna (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HIVE-16807?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mala Chikka Kempanna updated HIVE-16807:
----------------------------------------
    Summary: Persist row and column stats for HDFS directory across multiple external tables  (was: Persist table and column stats for HDFS directory across multiple external tables)

> Persist row and column stats for HDFS directory across multiple external tables
> -------------------------------------------------------------------------------
>
>                 Key: HIVE-16807
>                 URL: https://issues.apache.org/jira/browse/HIVE-16807
>             Project: Hive
>          Issue Type: New Feature
>            Reporter: Mala Chikka Kempanna
>
> HMS stores table and column stats for table which are tightly coupled to table names/ids.
> Instead, these computed stats should be tied to HDFS directory, so that even any external table created should be able to reuse and take advantage of previously computed stats. This conserves compute resources as there will be no need to expend them doing same work on same HDFS directory just for a different table name string



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)