Restarting and reinstating nodes
This section covers:
Stopping and restarting PAM Cluster nodes
Nodes within a PAM Cluster can be stopped at anytime but before stopping ensure users are logged off and there are no open device connections or device tasks running. If a node needs to be taken off line then use virtual appliance level Shut Down to allow for work in progress to be considered before shutting down. If you use the virtual appliance level Power Off there is a risk that corruption may occur.
When a node is stopped or fails the rest of the PAM Cluster will continue to operate but users logging onto the node will be disconnected and will be required to log onto a different node. When the node is started up again, the data will be automatically synchronised and the node will catchup with the rest.
If all nodes in a cluster go down due to power outage then when power has been restored, you can simply power them all on at once.
Reinstating a failed node
The method used to reinstate a failed node will depend on whether it is a leader or follower as well as the backup strategy implemented. For details on which restore procedure to use click here: Restoring clustered deployments.