openjpa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Abe White <>
Subject Re: Sequence configuration problems
Date Tue, 31 Oct 2006 22:42:48 GMT
> I investigated a bit further and found that if I change  
> PersistenceProviderImpl.ClassTransformerImpl to use a  
> JDBCConfigurationImpl instead of OpenJPAConfigurationImpl then the  
> problem goes away.  I'm reasonably certain that there is no need to  
> use a JDBCConfigurationImpl here in the class transformer, the  
> enhancing shouldn't need to know what the persistence mechanism  
> is.  Is it possible that the OpenJPAConfigurationImpl can ignore  
> the properties it doesn't understand rather than objecting to them?
> I've opened to  
> track this problem, that has a little more detail.  Given some  
> hints I can work towards a solution but I'm not sure where the best  
> place to start is.
> I'd really like to get this fixed ASAP in openjpa, we are hoping to  
> include a jpa/openjpa preview in the next geronimo release, and  
> this is the only significant problem I've encountered so far.  I've  
> successfully converted the Daytrader performance test app to  
> container managed jpa and it seems to work fine with openjpa.  Next  
> step there is to get some performance numbers :-)

I believe everything will work if you just remove the call to  
conf.instantiateAll() in the ClassTransformerImpl's constructor.   
It's not needed since the enhancement process doesn't require any  
instantiated plugins beyond the MetaDataRepository, which is already  
instantiated in the constructor.  (And I'm not even sure whether the  
ClassTransformer has to be threadsafe anyway -- the only reason we  
usually use instantiateAll() is to do up-front instantiation of  
singleton plugins to avoid locking and avoid lazy-instantiation  
threading issues.)

View raw message