Company1 originally set up template-based organization virtual datacenters that would all provide the same consumer offering within each available model as shown in the following table.

 

Pay-As-You-Go is an “all you can eat” model with the only restriction being the amount of resources available in the provider virtual datacenter. The settings above are assigned individually on a per virtual machine basis. In effect, the memory and CPU guarantees allocate physical resources to these virtual machines. Therefore, a 100% guarantee means that each virtual machine is granted 100% of its requested resources within the virtual datacenter, as is represented in the case of memory for High as long as the physical resources are available. In this example, a 2.5GHz CPU with 50% CPU and 100% memory guarantee could effectively be the highest performing virtual machine possible.

With the reservation pool model, the responsibility is on the consumer to assign physical resources from the pool to the virtual machines. This model effectively reserves 100% of the physical resources configured for CPU and memory from Company1’s provider virtual datacenter. CPU and memory resources cannot be used by other customers. Consumers do not necessarily get better performance, but there is the potential for higher performing and lower performing virtual machines within the consumer’s organization. The consumer must decide which virtual machines will be configured for higher or lower performance within the given pool of resources.Company1 has provided resources that the consumer can allocate however is desired. The levels represent the larger amounts of resources a customer has available. This increases the potential for more individual virtual machines to perform better, provided they are configured correctly.

 

 

With the Allocation Pool, the available resources in the pool are determined by the guarantees assigned. Listed above are the actual available resources within the pool for the consumer to use. This is similar to the Reservation Pool model. Effectively the high performance option is truly that, as it gives the pool a 100% guarantee not only on the resource pool, but also the individual virtual machines. The High option is almost a full combination of the other two models. The only real difference is when Low or Medium is used, some virtual machines can access the extra, unreserved resources depending on the configuration.