[
https://issues.apache.org/jira/browse/HIVE-3746?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13601584#comment-13601584
]
Phil Prudich commented on HIVE-3746:
------------------------------------
bq. The client is able to specify the maximum number of rows that should be
returned in each fetch request.
Right -- that's specifically the case I'm concerned with. If an application
has requested a single row, and the client has requested _n_ rows from the
server in an effort to reduce round trips, then _n_-1 intervening values from
the first column must be cached off somewhere before the first value for the
second column can be accessed.
Trying to address this by setting the row limit to 1 would also negatively
impact performance by requiring a round trip per row.
Without an intimate knowledge of the server code, I could see row orientation
as an optional argument to TFetchResultsReq, but would encourage that the
row-oriented behavior remain an option.
> TRowSet resultset structure should be column-oriented
> -----------------------------------------------------
>
> Key: HIVE-3746
> URL: https://issues.apache.org/jira/browse/HIVE-3746
> Project: Hive
> Issue Type: Sub-task
> Components: Server Infrastructure
> Reporter: Carl Steinbach
> Assignee: Carl Steinbach
> Labels: HiveServer2
>
--
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
