[
https://hibernate.onjira.com/browse/HHH-7206?page=com.atlassian.jira.plug...
]
Steve Ebersole commented on HHH-7206:
-------------------------------------
Another of the issues with long conversations (any time you have large numbers of stuff in
the persistence context really) is the amount of time needed to process dirtiness
calculation related to flushes. And here you are basically doing a throw away dirty
calculation. You calculate the dirtiness and if dirty you update the natural-id xrefs but
then just throw away the calculation. This is the concern I was addressing. And you can
in fact write just the natural-id changes to the database. The other changes you
mentioned would be discovered by dirtiness later and written during flush.
To be honest, I would never handle natural-id mutation in-line with other use-cases.
Thats asking for trouble IMO. So this use-case is foreign to me. Its ok with me, I
guess, to not do the write. As long as you realize the stylistic choice you are making
and understand the performance impact its really up to you as the app developer. As long
as Hibernate maintains correctness its really not important.
NaturalIdLoadAccess: NaturalId synchronization with no need to flush
at all
---------------------------------------------------------------------------
Key: HHH-7206
URL:
https://hibernate.onjira.com/browse/HHH-7206
Project: Hibernate ORM
Issue Type: Improvement
Components: core
Affects Versions: 4.1.1
Environment: Hibernate4, database-independent
Reporter: Guenther Demetz
Labels: flush, naturalId
With few code changes I think it is possible to make NaturalIdLoadAccess work with no
need to rely on flush at all,
please see following comments.
--
This message is automatically generated by JIRA.
For more information on JIRA, see:
http://www.atlassian.com/software/jira