Before You Begin
Consult the documentation for the resource type to determine what you must do before installing the upgrade package on a node. One action from the following list will be required:
You must reboot the node in noncluster mode.
You can leave the node running in cluster mode, but you must turn off monitoring of all instances of the resource type.
You can leave the node running in cluster mode and leave monitoring turned on for all instances of the resource type.
If you must reboot the node in noncluster mode, prevent a loss of service by performing a rolling upgrade. In a rolling upgrade, you install the package on each node individually while leaving the remaining nodes running in cluster mode.
To ensure that the correct version of the resource type is registered, you must specify the following information:
The resource type name
The resource type registration (RTR) file that defines the resource type
# clresourcetype register -f path-to-new-rtr-file resource-type-name
The format of the resource type name is as follows:
vendor-id.base-rt-name:rt-version
For an explanation of this format, see Format of Resource Type Names.
# clresourcetype show resource-type-name
You must perform this step if the package for the resource type upgrade is not installed on all cluster nodes.
The nodelist property of all resource groups that contain instances of the resource type must be a subset of the Installed_nodes property of the resource type.
# clresourcetype set -n installed-node-list resource-type
Specifies the names of nodes on which this resource type is installed.