Businesses today require more uptime than ever from their systems. The goal is to provide business continuity which also encompasses local high availability as well as disaster recovery. Business continuity targets must be agreed upon and documented as they guide the solutions that are deployed. Examples include overall service level agreements (SLAs), recovery time objectives (RTOs), and recovery point objectives (RPOs). Requirements must also align with the budget available which also must account for licensing.
There are two kinds of downtime: planned and unplanned. Unplanned downtime is exactly what it sounds like: unexpected events such as operating system issues, application-level errors, or hardware failures can spring up at any time with little to no warning causing an outage. No matter how brief, downtime still impacts the business. Planned downtime often occurs due to routine operations such as physical server, operating system, or other periodic maintenance tasks that require some level of outages. An example would be patching or updating a Windows Server host that would then need a reboot to complete the task.
Everything must be accounted for including the underlying infrastructure, the server and the hardware within it, Windows Server and Hyper-V, as well as the VMs and what is deployed in them. Reducing the impact of planned or unplanned outages can only happen if all layers are taken into account. For example, a networking outage will ultimately affect SQL Server running in a VM. It may appear to end users as a SQL Server outage, but the root cause is not SQL.