[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [eclipselink-users] using another L2 cache implementation
|
Hi Yannick,
A colleague pointed out to me that you might be able to benefit from
EclipseLink's cache coordination feature - and avoid all this work.
Cache Coordination info here:
http://wiki.eclipse.org/Introduction_to_Cache_%28ELUG%29#Cache_Coordination
Some Cache config info here including cache coordination:
http://wiki.eclipse.org/Using_EclipseLink_JPA_Extensions_%28ELUG%29#Using_EclipseLink_JPA_Extensions_for_Entity_Caching
Is that a possibility?
-Tom
Yannick Majoros wrote:
Le 19/07/2011 15:49, Tom Ware a écrit :
Hi Yannick,
I've been trying to track down an example for you to look at, but
unfortunately there is not one in the open source product.
The idea is that you subclass CacheInterceptor. Unfortunately, you
will have to override pretty much every method if you are completely
replacing the cache. We have been trying to figure out how to make
this it easier to plug in another cache, but unfortunately that work
is still in the discussion stages.
CacheInterceptor is designed to replace the identity map, and there
is no reason you need the target identity map. You will be replacing
every reference to targetIdentityMap in the CacheInterceptor code with
a call into your caching solution.
I appologize that is code is not easily overridable at the moment.
Hopefully in a future version we'll have a more user-friendly solution.
Hi Tom,
Thanks for your answer.
I'm still trying to decide if we will do that, or just use EL cache
coordination on a Glassfish cluster.
Any idea on performance? Is cache coordination a bottleneck? Should I
make the cache quite big and just hope many queries run against it?
Regards,
Yannick
_______________________________________________
eclipselink-users mailing list
eclipselink-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/eclipselink-users