Set up an IBM i database in a cluster
To communicate with a database, servers running IBM i can use either of two JDBC drivers: the IBM Toolbox for Java JDBC driver or the IBM Developer Kit for Java JDBC driver (also referred to as the native JDBC driver). Which JDBC driver use depends on how we are setting up the clustered environment.
The JDBC driver is specified by the db2_iseries.DbDriver property in wkplc_dbtype.properties, located in theWP_PROFILE/ConfigEngine/properties directory. We can specify the value by editing the file manually or by selecting the appropriate value using the configuration wizard.
- Native JDBC driver: com.ibm.db2.jdbc.app.DB2Driver
- IBM Toolbox for Java JDBC driver: com.ibm.as400.access.AS400JDBCDriver
Scaling topology considerations
Vertical and horizontal scaling topologies in an IBM i environment require different JDBC driver configurations, according to how you deploy the database.
| Scaling topology | JDBC driver considerations |
|---|---|
| Vertical scaling | When setting up a vertical cluster, we can install the database locally on the same machine as the portal or remotely on a separate machine. Use the appropriate JDBC driver, depending on where the database is installed.
|
| Horizontal scaling | When setting up a horizontal cluster, use the IBM Toolbox for Java JDBC driver. The typical configuration is to use a remote database for primary and secondary nodes in the cluster. If you choose, we can use a local database for the primary node and configure the secondary nodes to use that database, just as you would any other remote database. However, regardless of whether you choose to include a local database in the environment, use the IBM Toolbox for Java JDBC driver with the horizontal cluster. |
Use a local database in an IBM i horizontal cluster
Although the instructions for setting up a horizontal cluster describe how to use a remote database for both primary and secondary nodes, we can choose to configure the IBM i horizontal cluster to use a local database for the primary node instead. In this example, a database and Web Server are locally installed on the system (System 1) where IBM WebSphere Portal and IBM WAS are installed. System 1 is the primary node. System 2 is the secondary node.
Although it is possible to use a local database on a secondary node instead of the primary node, this configuration has not been tested and is not documented here.
Important: Even though we are using a local database for the primary node in this scenario, all database connections are configured as if the database were remote. Specifically, use the IBM Toolbox for Java JDBC driver (com.ibm.as400.access.AS400JDBCDriver) when configuring the database for both primary and secondary nodes.
To use a local database with the primary node, perform the database configuration, with the following variations when updating properties files in theWP_PROFILE/ConfigEngine.
- wkplc_dbtype.properties
-
- Specify the JDBC driver in the db2_iseries.DbDriver property.
For example:
db2_iseries.DbDriver=com.ibm.as400.access.AS400JDBCDriver
- Specify the database location as remote in the db2_iseries.DbDriverType property.
For example:
db2_iseries.DbDriverType=4
- Specify the JDBC driver in the db2_iseries.DbDriver property.
- wkplc_comp.properties
-
- Specify the primary node's host name for the domain.DbName properties.
For example: release.DbName=primary_host_name/wpsdb
- Specify the primary node's host name in the domain.DbUrl properties.
For example: release.DbUrl=jdbc:as400:primary_host_name/wpsdb
- Specify the primary node's host name for the domain.DbName properties.
If using the configuration wizard for database transfer, update the values in the wizard panels rather than in the properties files.
Complete all other configuration as described. When configuring secondary nodes in this scenario, perform the database configuration as you would for any remote database, using the primary node's host name for the database transfer.
Parent: Cluster considerations