What is a hypervisor?
A hypervisor is a software layer that allocates one physical computer’s resources among virtual machines.
How it works and where it is used
Type 1 hypervisors run close to hardware; Type 2 run atop a host OS, though products can be more complex. Each VM has its own OS. Overcommitting memory, CPU or storage can reduce performance, requiring reservations and measurement. VMs and containers use different isolation models.
Limits, safety and selection criteria
Migration can ease maintenance but shared storage and network failures remain risks. Restrict management access, patch the hypervisor and update guest systems separately. Backup images are not trustworthy without restore tests. Two VMs on one physical server may run separate operating systems yet share electricity, network interfaces and storage. One VM’s heavy disk use can affect another. A snapshot can support short-term rollback, but long-lived snapshots consume storage and are not substitutes for backups. The management network can be separated from workload traffic. Capacity planning should test not only normal load but how remaining hosts handle a physical-node failure. Isolation reduces some risks, yet privileged access to the host remains especially sensitive.