|
|
Failover with ARM and PRR is a cold failover. With the introduction of the WebSphere HAManager in WAS V6, you have the option of hot standby and peer recovery. It does not restart a server. Instead, the transaction on the failing server is recovered by a different appserver in the same cluster. For more information about HAManager and peer recovery, see Chapter 6, WebSphere HAManager.
Only the Type 4 JDBC driver is supported by the HAManager. HAManager is required for Web Services, the default messaging provider, JCA 1.5 connectors and so on. It is recommended to gradually migrate to the HAManager approach because the PRR function will not be supported in future releases of WAS on z/OS.
|
Important: Use either HAManager or PRR for recovery in WebSphere V6. However, do not use both, because unpredictable results can occur. |
