Details on incompatibilities 1
At the time of forced startup of a cluster application, if any node where RMS is not running exists, starting the cluster application as it is causes the risk of data corruption.
Therefore, to deal with the issue, the function is added. The function forcibly shuts down any node where RMS is not running before forcibly starting the cluster application.
Changes
At the time of forced startup of a cluster application, even if any node where RMS is not running exists in the cluster, which may result in data corruption, the cluster application is forcibly started according to the user's operation.
For reducing the risk of data corruption at the time of forced startup of a cluster application, the function forcibly starts the cluster application after forcibly shutting down any node where RMS is not running.
Note
Read through the items in "4.2 Operator Intervention Messages" in "PRIMECLUSTER Messages."
Details on incompatibilities 2
With the default settings made when the cluster was installed, the operator intervention request is always enabled. For details, see "5.4 Setting Up Fault Resource Identification and Operator Intervention Request."
Changes
The operator intervention request will not work with the default setting at installation. The default value of AppWatch set when the cluster was installed is set to OFF, and the operator intervention request will not work with this default value.
The operator intervention request will work with the default setting at installation. The operator intervention request, is disabled only when the AppWatch parameter is set to OFF with clsetparam.
Note
After you have changed the AppWatch parameter with clsetparam, you have to restart all the nodes to validate the setting.
