Consider the following scenario:
- You have nodes that are running on a Windows Server 2012 R2-based failover cluster.
- There are many disks in the storage space pool.
- One of the nodes fails.