Has anyone tried to use Hibernate with C-JDBC? (http://c-jdbc.objectweb.org/) As hibernate doesn't use (or encourage) stored procedures, it seems like it could be very effectively.

Specifically what I'm looking to do is use C-JDBC to a local hsql database, and use C-JDBC to merge that with a production database. Then all that static data like user types and other things can exist in a virtual database and be accessed invisibly using Hibernate.

Just curious if anybody has used these two or if any of this sounds like there will be unforeseen problems.

I've been wanting to check out C-JDBC for quite a while, but something else always comes up :)

So you want to do something like store all the "reference" data in a local hypersonic database and the rest of the data housed in a seperate production DB and then use C-JDBC to mediate access between the two DBs?

Should be OK. The only issue I could forsee is an inability to defiine proper constraints.

I actually want to look at C-JDBC for its load-balancing support across multiple database servers. Oracle itself has something similiar and we may end up going with that, however. That has the potential to be such a huge performance booster.

We too have been following C-JDBC progress for the clustering support. One idea had been to keep our primary DB server MSSQL, but then add some extra MySQL database servers that can share the read-load.

But this idea is to be able to keep all the static tables in CVS with the Java code, but still access it as part of the existing database (enforcing primary keys and accessed via Hibernate).

C-JDBC allows you to create layers, so roughly our architecture looks like this:

Layer 1 is RAIDb 0 (tables across databases), splitting static tables into HSQL and regular tables into Layer 2.

Layer 2 is RAIDb 1 (mirrored across databases), having 1 MSSQL and n MySQL databases mirroring the same tables.

C-JDBC supposedly let's you treat the underlying databases agnostically, but I'm not sure how that will work. Hibernate does that to a large extent, but it has well defined code (dialects) to map features to underlying databases. Not sure how those layers will work. Might need to create a C-JDBC dialect?

Just not sure. If you do get around to testing that, let me know and vice versa.

I think it is important to note that distributed join is not supported by c-jdbc. So, queries on distributed table would not work.


6.7. Current Limitations
The C-JDBC controller in its 1.1 release has the following limitations:
• transparent controller failover is not supported when using horizontal scalability with JGroups,
• network partition/reconciliation is not supported,
• distributed joins are not supported which means that you must ensure that every query
least a single backend,
• RAIDb-1ec and RAIDb-2ec levels are not supported,
• the FileRecoveryLog is not yet implemented, use the JDBCRecoveryLog instead.

(c-jdbc 1.1 user guide, p. 27)

etienne

did you guys ever get the time to experient hibernate with C-JDBC? Anybody else?

I am seriously considering using c-jdbc with hibernate and would like to know you guys opinion on that...