Routine cluster maintenance implies installing updates, replacing failed disks, backing up and restoring the management database. The cluster maintenance also includes gracefully releasing nodes from the storage cluster without data loss and removing them from the infrastructure. You can perform node maintenance by placing a node into the maintenance mode. When changing cluster nodes, you might need to modify the high availability configuration or even destroy and re-create it. Additionally, it is important to know the recommended way to gracefully shut down all nodes and start up the cluster later.