I am on a project that was started (not by me) years ago. Today, the
project still uses Oracle9iASTopLink. I found out about EclipseLink and am
pondering an upgrade. However, without "valid reason" the client will not
pay for my time to actually do an upgrade.
So, in terms of sheer performance, will I see an immediate gain in TopLink
performance? Is the architecture of the code so different now that I
should upgrade?
Basically I am looking for links to other resources that I can read so I
can understand the implications of upgrading. But from you guys, I was
wondering what your general feeling on the matter is.
Are you of the opinion that I would be able to notice immediate
improvements in performance (our app is essentially a search engine) just
by moving from the older Toplink to EclipseLink? Or should I not be
expecting that.
Yes, that is true. However, talking with a friend of mine, he was
mentioning that there was a lot of change to the core programming of
EclipseLink compared to what code is executing in my current version of
TopLink. If significant leaps forward have been made in terms of cutting
down processing or execution time when TopLink is interacting with the
DataBase, we figured there might just be a performance improvement right
out of the box so to speak.
Now I know performance is a complicated subject - but I thought Id start a
thread brewing...and see if I could learn anything.
We contributed the entire TopLink code-base to EclipseLink. The only
differences you will find beyond the package renaming and removal of
deprecated code should be the removal of some Oracle specific dependencies
on our container and significant new functionality around JPA, JAXB2, and
SDO.
I would like to also mentioned that there were significant performance
improvments (specifcially to concurrency) between TopLink 9 and TopLink
11/EclipseLink and we are committed to improving performance on an ongoing
basis.
It should be a relatively easy switch over for you as Doug mentioned. Try
it out and see for yourself.
--Gordon