Sure thing. I spent the time to put some debug code in-place and figure out the problem. In PersistenceUnitBuilder.initContext(), there was no condition being put in place for the the fact that the moduleContext.getBaseDir() is not different regardless of whether or not the ModuleContext was constructed with inPlaceConfiguration dir not being null. Rather than changing the ModuleContext to help support a boolean for this value (which I think probably might have been cleaner), I went a layer up and checked to see if the configurationData of the rootEar had a null inPlaceConfigurationDir which would indicate the same condition from above. I adjusted the rootBaseFile and moduleBaseURI accordingly if it wasn't null from the when the MoudleContext was built.
I've created a JIRA issue for it located here: https://issues.apache.org/jira/browse/GERONIMO-5452 In addition, I've attached the fix here that I wrote for the 2.1.6 branch. http://apache-geronimo.328035.n3.nabble.com/file/n975531/inplace-persistencebuilder.patch inplace-persistencebuilder.patch Thanks for allowing me the opportunity to contribute. -- View this message in context: http://apache-geronimo.328035.n3.nabble.com/inplace-deployment-and-persistence-xml-tp965725p975531.html Sent from the Users mailing list archive at Nabble.com.
