Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.java.programmer > #33151 > unrolled thread
| Started by | Cecil Westerhof <Cecil@decebal.nl> |
|---|---|
| First post | 2015-11-20 09:41 +0100 |
| Last post | 2015-11-20 19:24 -0500 |
| Articles | 2 — 2 participants |
Back to article view | Back to comp.lang.java.programmer
Performance JPA Cecil Westerhof <Cecil@decebal.nl> - 2015-11-20 09:41 +0100
Re: Performance JPA Arne Vajhøj <arne@vajhoej.dk> - 2015-11-20 19:24 -0500
| From | Cecil Westerhof <Cecil@decebal.nl> |
|---|---|
| Date | 2015-11-20 09:41 +0100 |
| Subject | Performance JPA |
| Message-ID | <87h9khysdn.fsf@Equus.decebal.nl> |
I am at the end of my first JPA project. At the moment we are working with one entity manager in the application. I heard that it could be a good idea to use several entity managers, because that could improve the performance. Is this true, or just a myth? Also I was told that Hibernate is about three times as fast as openJPA. We are required to use openJPA because we are using the IBM Web Application Server. The performance with openJPA is about the same as when the application used JDBC. So then it looks not very reasonable to think that Hibernate will be three times as fast as openJPA. But I also understood that IBM WAS uses a highly optimised version of openJPA. Is it true that Hibernate is a lot faster, or is it a case of: lies, damned lies and statistics? -- Cecil Westerhof Senior Software Engineer LinkedIn: http://www.linkedin.com/in/cecilwesterhof
[toc] | [next] | [standalone]
| From | Arne Vajhøj <arne@vajhoej.dk> |
|---|---|
| Date | 2015-11-20 19:24 -0500 |
| Message-ID | <n2odjb$ghe$1@speranza.aioe.org> |
| In reply to | #33151 |
On 11/20/2015 3:41 AM, Cecil Westerhof wrote: > I am at the end of my first JPA project. At the moment we are working > with one entity manager in the application. I heard that it could be a > good idea to use several entity managers, because that could improve > the performance. Is this true, or just a myth? > > > Also I was told that Hibernate is about three times as fast as > openJPA. We are required to use openJPA because we are using the IBM > Web Application Server. The performance with openJPA is about the same > as when the application used JDBC. So then it looks not very > reasonable to think that Hibernate will be three times as fast as > openJPA. But I also understood that IBM WAS uses a highly optimised > version of openJPA. > > Is it true that Hibernate is a lot faster, or is it a case of: lies, > damned lies and statistics? Damn lies. :-) For any persistence framework the majority of time should be spent on database work and communication with the database. If OpenJPA already send good SQL, then it can only be improved marginally. Furthermore JPA providers is a competitive area, so if one provider got a brilliant idea, then the other would adopt it immediately. Arne
[toc] | [prev] | [standalone]
Back to top | Article view | comp.lang.java.programmer
csiph-web