[
https://issues.apache.org/jira/browse/SQOOP-1701?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14204460#comment-14204460
]
Jarek Jarcec Cecho commented on SQOOP-1701:
-------------------------------------------
Thank you for putting the note together [~daisuke.kobayashi]. Couple of notes:
* Let's move the note from MySQL Direct connector into MySQL JDBC Connector
section as that is where this parameter is applicable.
* I would suggest to easy the language a bit as it's not entirely correct that
the argument {{--fetch-size}} will be ignored. User can still use it with the
cave-head that only two modes are supported.
> MySQL fetch-size behavior should be documented explicitly
> ---------------------------------------------------------
>
> Key: SQOOP-1701
> URL: https://issues.apache.org/jira/browse/SQOOP-1701
> Project: Sqoop
> Issue Type: Improvement
> Components: docs
> Affects Versions: 1.4.5
> Reporter: Daisuke Kobayashi
> Attachments: SQOOP-1701.patch
>
>
> The behavior of {{--fetch-size}} has been changed in SQOOP-1400 and reverted
> in SQOOP-1617 as follows:
> - Prior to SQOOP-1400: row-by-row is the default
> - With SQOOP-1400: load everything is the default
> - SQOOP-1617 and higher: row-by-row becomes the default again
> The behavior of the option with MySQL is a bit different from the other
> drivers because MySQL JDBC driver can only handle row-by-row or
> load-everything. So that it should be documented in the area of the
> limitation.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
