I think I found what I was searching for: http://openjpa.apache.org/builds/2.2.1/apache-openjpa/docs/ref_guide_event.html#ref_guide_event_conf_tcp
The RemoteCommitProvider must be set up and notify the other instances to have a cache synchronization. I am not sure about the primary key issue yet, I guess there must be some kind of solution to let the database-cluster manage the sequences/pk's, so that multiple instances can write to the same database-cluster at once without primary key duplication/constraint errors. Thanks Sebastian 2013/1/15 seba.wag...@gmail.com <seba.wag...@gmail.com> > Hi, > > we are currently drafting/implementing some clustering mechanisms for > Apache OpenMeetings. > > I have seen OpenJPA supports the MySQL Replication feature: > > http://people.apache.org/~mikedd/nightly.builds/apache-openjpa-2.3.0-SNAPSHOT/docs/docbook/manual/ref_guide_dbsetup_dbsupport.html#ref_guide_dbsetup_dbsupport_mysql > which is great! > > However it solves only one side of the cluster: Having the database > clustered. > > What happens if you want to cluster OpenJPA itself on multiple nodes? Each > OpenJPA instance would have its own data-cache, how can you make sure those > caches are synchronized across all nodes in your cluster? > And how would you solve the primary key/sequencer issue? When multiple > instances write to the same database they could potentially try to create a > record with the same primary key. > > Thanks! > Sebastian > > -- > Sebastian Wagner > https://twitter.com/#!/dead_lock > http://www.webbase-design.de > http://www.wagner-sebastian.com > seba.wag...@gmail.com > -- Sebastian Wagner https://twitter.com/#!/dead_lock http://www.webbase-design.de http://www.wagner-sebastian.com seba.wag...@gmail.com