It would depend on what is inside the CacheKey, which is the key object Hibernate passes to the second-level cache. Usually this situation comes up with the identifier Type(s), especially in the case of composite identifiers. Some of the Type implementations hold reference to the SessionFactory. Serializing those properly depends on being able to resolve the proper SessionFactory reference on deser, which can happen based on UUID (same JVM) or name (across JVMs).
On Mon, Nov 10, 2014 at 4:17 AM, Galder Zamarreño <gal...@redhat.com> wrote: > Hi all, > > Re: > http://stackoverflow.com/questions/26110918/infinispan-marshalling-error-for-hibernate-entity-with-composite-primary-key > Re: https://github.com/hibernate/hibernate-orm/pull/820 > > Seems like the issue is resolved by making sure Session Factories have the > same name. > > Is this something new? I’ve never had the need to set it in order to get > Hibernate 2LC clustered tests working. None of my unit tests set the > session factory name. One of my most recent demos uses JPA, and never had > to set hibernate.session_factory_name property. > > Or is this due to the peculiarity of the Entity the user has? I’ve run > some tests locally and SessionFactoryImpl.readResolve() does not seem to > get called. > > @Paul/@Scott, have you seen anything similar before? > > Cheers, > -- > Galder Zamarreño > gal...@redhat.com > twitter.com/galderz > > > _______________________________________________ > hibernate-dev mailing list > hibernate-dev@lists.jboss.org > https://lists.jboss.org/mailman/listinfo/hibernate-dev > _______________________________________________ hibernate-dev mailing list hibernate-dev@lists.jboss.org https://lists.jboss.org/mailman/listinfo/hibernate-dev
