[jira] [Commented] (HIVE-18328) Improve schematool validator to report duplicate rows for column statistics

["Aihua Xu (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-18328?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16354796#comment-16354796 ] 

Aihua Xu commented on HIVE-18328:
---------------------------------

[~ngangam] Sorry that I missed this one. Do you think we can also add similar validation for PART_COL_STATS as well? 

> Improve schematool validator to report duplicate rows for column statistics
> ---------------------------------------------------------------------------
>
>                 Key: HIVE-18328
>                 URL: https://issues.apache.org/jira/browse/HIVE-18328
>             Project: Hive
>          Issue Type: Improvement
>          Components: Hive
>    Affects Versions: 2.1.1
>            Reporter: Naveen Gangam
>            Assignee: Naveen Gangam
>            Priority: Major
>         Attachments: HIVE-18328.patch
>
>
> By design, in the {{TAB_COL_STATS}} table of the HMS schema, there should be ONE AND ONLY ONE row, representing its statistics, for each column defined in hive. A combination of DB_NAME, TABLE_NAME and COLUMN_NAME constitute a primary key/unique row.
> Each time the statistics are computed for a column, this row is updated. 
> However, if somehow via  BDR/replication process, we end up with multiple rows in this table for a given column, HMS server to recompute the statistics there after.
> So it would be good to detect this data anamoly via the schema validation tool.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)