[hibernate-issues] [Hibernate-JIRA] Commented: (HHH-3011) merge() optimization for Oracle dialect

[Emmanuel Bernard (JIRA)]

    [ http://opensource.atlassian.com/projects/hibernate/browse/HHH-3011?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_29145 ] 

Emmanuel Bernard commented on HHH-3011:
---------------------------------------

this is probably not equivalent because a trigger can be raised at the db level in the second case and not the first.
Plus depending on the graph you need to merge, the hibernate impl will be faster than the SQL MERGE, essentially limited to 1 select query for n different "merges" if no change is detected.

> merge() optimization for Oracle dialect
> ---------------------------------------
>
>                 Key: HHH-3011
>                 URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-3011
>             Project: Hibernate3
>          Issue Type: Improvement
>          Components: core
>            Reporter: MG
>            Priority: Minor
>
> merge() optimization for Oracle dialect.
> Since 9i Oracle has a MERGE statement which is very useful to save round trips to database in merge() call
> which will make dramatic improvement for performance  of loading large row sets.
> MERGE command processes the conditional INSERT-or-UPDATE within a single statement.
> So instead of doing 'select' first to detect whether entity is in database or not (This is one round trip to DB)
> and then issuing 'update' or 'insert' all of it could be done in one statement with no extra round trips using oracle SQL MERGE statement.
> http://www.oracle.com/technology/products/oracle9i/daily/Aug24.html

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira