HA Data source... how?

[pt00dgr]

Hi. I am clustering for high availability which means that I want to eliminate the single points of failure in my system. The clustering works just fine but I still have one single point of failure, ie. the data source (MySql).

My questions:
Does Jboss support redundant data bases or do I need to take care of that with other techniques?
Where can I read about this?

Thanks in advance

/David

[ivelin.ivanov]

What I usually see people do is run their database on a RAID array. This way you have an HA data store and therefore an HA middle tier. Cheap and working solution.

Alternatively, MySQL is working on an HA version which distributes the load and data across the cluster nodes. Oracle and DB2 have HA offerings for a long time.


Ivelin

[nickman]

You might also look at ObjectWeb's Clustered-JDBC. Seems pretty decent, but I have not used it.
C-JDBC: http://c-jdbc.objectweb.org/