hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Nemon Lou (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HIVE-12382) return actual row count for JDBC executeUpdate
Date Tue, 24 Nov 2015 13:09:11 GMT

    [ https://issues.apache.org/jira/browse/HIVE-12382?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15024468#comment-15024468
] 

Nemon Lou commented on HIVE-12382:
----------------------------------

Design proposal:
1,Adding a new counters "RECORD_OUT" in FileSinkOperator
   JDBC executeUpdate only returns one int result ,so we only need one result for all the
output tables.(multiple insert ,for example)
2,Using the existing FetchResults thirft API for fetching affected row count by adding a new
fetch type.

Only MapReduce and Spark are supported. It's easy to support Tez .


> return actual row count for JDBC executeUpdate
> ----------------------------------------------
>
>                 Key: HIVE-12382
>                 URL: https://issues.apache.org/jira/browse/HIVE-12382
>             Project: Hive
>          Issue Type: Improvement
>          Components: JDBC
>            Reporter: Nemon Lou
>            Assignee: Nemon Lou
>            Priority: Minor
>         Attachments: HIVE-12382.patch
>
>
> when running sql like 'insert into/overwrite table',
> user may want to know how many rows are inserted .
> Return actual row count for HiveStatement.executeUpdate is useful in such cases.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message