Just some general notes in case someone wants to attempt a port: Do the facilitating refactorings first (e.g. centralising cache key creation) without new features and submit that. Then add the features in separate pull requests. Much easier to review and understand that way (read: speedier).
/Oskar 2012/3/13 Gerke Geurts <[email protected]>: > Based on a fairly quick scan of the Hibernate multitentant changes, I would > say that porting should be feasible enough that I would attempt to do it > myself if we hadn't already spent time already on getting NHibernate Shards > into sufficient shape for a similar purpose. > > The main areas affected seem to be: > > Extension of connection provider infrastructure to enable use of single > session factory with multiple databases > Modification of second-level cache code to enable inclusion of tenant > identifier in cache keys. This is achieved by centralising cache key > creation in the session. > Modification of persistence context code to enable inclusion of tenant > identifier in entity keys. This is achieved by centralising entity key > creation in the session. > Extension of configuration code to specify tenant strategy. > Extension of session interface to specify to tenant identifier for a > session. > > Regards, > Gerke. > > > On Tuesday, 13 March 2012 17:44:42 UTC+1, Paul White wrote: >> >> Hello, >> >> I'm new to this group and have a few questions regarding Hibernate <--> >> NHibernate >> >> Are the two projects still close enough to consider porting features? >> >> Do new Hibernate features generally get ported to NHibernate and is there >> any plan for porting the recent changes in Hibernate (4, 4.1)? >> >> Can I help? I'd like to contribute if I can with respect to multi-tenancy >> in particular. >> >> Documentation of multi-tenancy support in Hibernate >> 4.1: http://docs.jboss.org/hibernate/core/4.1/devguide/en-US/html/ch16.html#d5e4617 >> >> Thanks, >> >> -Paul
