[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Paul Bakker updated HHH-5303: ----------------------------- Attachment: testcase.zip -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Paul Bakker commented on HHH-5303: ---------------------------------- I have the following in my persistence.xml: <property name="javax.persistence.sharedCache.mode" value="ENABLE_SELECTIVE"/> Also tried "ALL". -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Paul Bakker commented on HHH-5303: ---------------------------------- I'm using EhCache. If I don't specify the concurrency setting (second property) nothing happens. If I do, I can't turn caching of using @Cacheable(false). {code} <property name="hibernate.cache.provider_class" value="org.hibernate.cache.EhCacheProvider"/> <property name="hibernate.ejb.classcache.demo.entities.Employee" value="read-write"/> {code} Looking at determineSharedCacheMode I don't see anything that could go wrong with my setup. Tomorrow I'll try to run a debug session and see if I find anything. Note that my project (just an example project demonstrating the problem) is attached too. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Steve Ebersole commented on HHH-5303: ------------------------------------- The only ways that I can see cacheAnn as null is that you did not specify @Cache and either: # You used SharedCacheMode.NONE or UNSPECIFIED (or did not specify) # You used SharedCacheMode.ENABLE_SELECTIVE and had @Cacheble(false) or did not specify @Cacheable # You used SharedCacheMode.DISABLE_SELECTIVE and had @Cacheble(false) -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Steve Ebersole commented on HHH-5303: ------------------------------------- Just a raw @Cacheable annotation is quite difficult for Hibernate to decipher because of its pluggable cache region factories and granular access types, especially because not all caching providers support all of the access types. As of this moment, what it does is to ask the RegionFactory (the "cache provider") which is the default. So as I asked a few posts ago, it depends on which factory/provider you have specified. Looking at org.hibernate.cfg.AnnotationBinder#prepareDefaultCacheConcurrencyStrategy (not currently used) it seems that either I planned on allowing setting this via config but forgot to "hook it up" or allowed it before adding the org.hibernate.cache.RegionFactory#getDefaultAccessType method and forgot to remove it. This was done at a time of a lot of rapid development for JPA2 and I did not do a great job of documenting the intentions here in the code. But i remember having a lot of discussion of this at the time. It seems reasonable right now to blend these 2 things together such that we first look for an explicit config value (ala #prepareDefaultCacheConcurrencyStrategy) and use that if defined; otherwise we look to the RegionFactory for it's default. I need to think that through a bit though. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira

[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-5303?page=c... ] Steve Ebersole commented on HHH-5303: ------------------------------------- Its not just difficult, its impossible. Why? Well say you chose READ_WRITE as the default as you suggest. The problem is that if you run with JBossCache or (I believe) Infinispan, READ_WRITE is not a valid option. The underlying problem is that (1) not all providers support all access types and (2) we do not really define a "hierarchy" amongst the access types to be able to switch over from one to the other automatically if a provider did not support the specified one. And its not just an issue with READ_WRITE itself as the specific access type. -- 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.... - For more information on JIRA, see: http://www.atlassian.com/software/jira