[
https://issues.apache.org/jira/browse/HIVE-3746?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504959#comment-13504959
]
Phil Prudich commented on HIVE-3746:
------------------------------------
To make sure I'm reading the new thrift definitions correctly -- does this mean
that all rows' column 1 values will come first on the wire, and then be
followed by all rows' values for column 2, and so on? I clearly see how this
would save bytes on the wire.
However, any client trying to return rows one-at-a-time to an application would
be required to read, process, and buffer almost an entire reply-worth of data
before being able to return the first complete row.
I'm unfamiliar with the server code; but similar buffering may be needed there
as well.
Is my understanding of the issue correct?
> 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
>
--
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
