How do you implement a clustered database?

[Jane Jukowsky]

My database server (h2) does not reliably support clustering yet; but I need to insure data redundancy. Can I mess up with Hibernate connections in such a way that every database connection would in fact be two connections to two different databases on two computers, seamlessly for the rest of the application? I only care about writes; reads can go to one database as far as I am concerned.

[Mark Spritzler]

Well you could use Hibernate Shards project possibly.

But with Hibernate a Connection factory is one to one to a SessionFactory. So to point to two different databases you need two different SessionFactories.

Mark

[Jane Jukowsky]

What if I supply my own ConnectionFactory, is that very convoluted?

[Mark Spritzler]

Jane Dodo wrote:What if I supply my own ConnectionFactory, is that very convoluted?

a single Hibernate SessionFactory can only communicate with a single database.

Mark