You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hive.apache.org by "Shreepadma Venugopalan (JIRA)" <ji...@apache.org> on 2012/11/15 20:33:12 UTC

[jira] [Created] (HIVE-3712) Use varbinary instead of longvarbinary to store min and max column values in column stats schema

Shreepadma Venugopalan created HIVE-3712:
--------------------------------------------

             Summary: Use varbinary instead of longvarbinary to store min and max column values in column stats schema
                 Key: HIVE-3712
                 URL: https://issues.apache.org/jira/browse/HIVE-3712
             Project: Hive
          Issue Type: Bug
          Components: Metastore, Statistics
    Affects Versions: 0.9.0
            Reporter: Shreepadma Venugopalan
            Assignee: Shreepadma Venugopalan


JDBC type longvarbinary maps to BLOB SQL type in some databases. Storing min and max column values for numeric types takes up 8 bytes and hence doesn't require a BLOB. Storing these values in a BLOB will impact performance without providing much benefits. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Updated] (HIVE-3712) Use varbinary instead of longvarbinary to store min and max column values in column stats schema

Posted by "Ashutosh Chauhan (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/HIVE-3712?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ashutosh Chauhan updated HIVE-3712:
-----------------------------------

       Resolution: Fixed
    Fix Version/s: 0.10.0
           Status: Resolved  (was: Patch Available)

Taken care of in HIVE-3678
                
> Use varbinary instead of longvarbinary to store min and max column values in column stats schema
> ------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-3712
>                 URL: https://issues.apache.org/jira/browse/HIVE-3712
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore, Statistics
>    Affects Versions: 0.9.0
>            Reporter: Shreepadma Venugopalan
>            Assignee: Shreepadma Venugopalan
>             Fix For: 0.10.0
>
>
> JDBC type longvarbinary maps to BLOB SQL type in some databases. Storing min and max column values for numeric types takes up 8 bytes and hence doesn't require a BLOB. Storing these values in a BLOB will impact performance without providing much benefits. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HIVE-3712) Use varbinary instead of longvarbinary to store min and max column values in column stats schema

Posted by "Shreepadma Venugopalan (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/HIVE-3712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13499924#comment-13499924 ] 

Shreepadma Venugopalan commented on HIVE-3712:
----------------------------------------------

It looks like VARBINARY is not supported across different DBs and DB versions in a consistent manner. Storing 8 bytes in a LONGVARBINARY is an overkill because the LONGVARBINARY is mapped to BLOB type in some DBs. It appears the best solution at this point is to store LONG and DOUBLE min and max values in two separate columns. 
                
> Use varbinary instead of longvarbinary to store min and max column values in column stats schema
> ------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-3712
>                 URL: https://issues.apache.org/jira/browse/HIVE-3712
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore, Statistics
>    Affects Versions: 0.9.0
>            Reporter: Shreepadma Venugopalan
>            Assignee: Shreepadma Venugopalan
>
> JDBC type longvarbinary maps to BLOB SQL type in some databases. Storing min and max column values for numeric types takes up 8 bytes and hence doesn't require a BLOB. Storing these values in a BLOB will impact performance without providing much benefits. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HIVE-3712) Use varbinary instead of longvarbinary to store min and max column values in column stats schema

Posted by "Shreepadma Venugopalan (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/HIVE-3712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13499931#comment-13499931 ] 

Shreepadma Venugopalan commented on HIVE-3712:
----------------------------------------------

Review board link: https://reviews.apache.org/r/8119/
                
> Use varbinary instead of longvarbinary to store min and max column values in column stats schema
> ------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-3712
>                 URL: https://issues.apache.org/jira/browse/HIVE-3712
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore, Statistics
>    Affects Versions: 0.9.0
>            Reporter: Shreepadma Venugopalan
>            Assignee: Shreepadma Venugopalan
>
> JDBC type longvarbinary maps to BLOB SQL type in some databases. Storing min and max column values for numeric types takes up 8 bytes and hence doesn't require a BLOB. Storing these values in a BLOB will impact performance without providing much benefits. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Updated] (HIVE-3712) Use varbinary instead of longvarbinary to store min and max column values in column stats schema

Posted by "Shreepadma Venugopalan (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/HIVE-3712?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Shreepadma Venugopalan updated HIVE-3712:
-----------------------------------------

    Status: Patch Available  (was: Open)
    
> Use varbinary instead of longvarbinary to store min and max column values in column stats schema
> ------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-3712
>                 URL: https://issues.apache.org/jira/browse/HIVE-3712
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore, Statistics
>    Affects Versions: 0.9.0
>            Reporter: Shreepadma Venugopalan
>            Assignee: Shreepadma Venugopalan
>
> JDBC type longvarbinary maps to BLOB SQL type in some databases. Storing min and max column values for numeric types takes up 8 bytes and hence doesn't require a BLOB. Storing these values in a BLOB will impact performance without providing much benefits. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira