Search code examples
weblogicweblogic12c

Are there any benefits to duplicate connection pools in Weblogic?

I have a cluster of 4 Weblogic instances on 2 servers. My JDBC connection pool is targeting the cluster.

Would there be any benefit if I duplicate the connection pool to the same database, and use the multi data source connection?

Is it possible that 1 connection pool could become unusable, but the other be ok, even though they are identical to each other? would this increase availability?

I tried to Google this, but could not find any answers.

Thanks, Bill

Solution

  • It depends on your database layer.

    1. One single database If you have one single database, there will be no benefit at having a Multi Data Source with two (or more) datasources pointing to the same database. If the database falls, there's no magic.

    2. Database Cluster If you have two databases in active-active mode, then the Multi Data Source feature will give you benefits. For this mode, choose Load Balancing algorithm: since both servers are active and with updated data then you may increase performance. If you have active-passive mode then you may consider Failover algorithm.

    For Oracle RAC, ignore Multi Data Source and focus on the GridLink Data Source.

    Make sure to read Configuring JDBC Multi Data Sources for 12.1.2.