Refer the following steps to design for scaling up or scaling out:
- Determine whether the host in the environment should scale up or scale out.
- Determine the number of virtual machine workloads per host.
- Based on the number of virtual machines per host, calculate the number of hosts that are required. This should also include the number of hosts required to support growth and failover:
(Number of Workloads / Number of Workloads per Host) + (Number of Future Workloads / Number of Workloads per Host) + Number of Failover Hosts = Number of Physical Hosts Required
- Using the identified CPU requirements, calculate the CPU resources required per host:
Total CPU Resources Required / (Number of Physical Hosts Required - Failover Hosts) = CPU Resources per Host
- Using the identified memory requirements, calculate the memory resources that are required per host:
Total Memory Resources Required / (Number of Physical Hosts Required - Failover Hosts) = Memory Resources per Host