As part of the design process for the Committed service offering, providers should pay special consideration to key service settings and values in vCloud Director that can affect service performance and consistency levels for a subscriber’s organization virtual datacenter. Despite use of the Allocation Pool allocation model in this service, circumstances may arise that can result in subscribers overcommitting resources over time. If not properly managed, performance for all application workloads can be negatively affected. Table 15 provides an example of these key service settings, values, and justifications.

 

In this example, the CPU allocation setting serves as a block or aggregate limit for the entire target organization virtual datacenter and has been configured as 50GHz (50,000 MHz). Of this 50GHz resource allocation, 75% (37.5GHz of total CPU capacity) is marked as guaranteed. This implies that the remaining 25% (12.5GHz of total CPU capacity) is available for bursting if sufficient infrastructure resources are available at the time of need.

The memory allocation setting also serves as a block, or aggregate, limit for the entire target organization virtual datacenter. It has been configured as double the CPU capacity, or 100GB. Of this 100GB resource allocation, 75%—75GB of total memory capacity—is marked as guaranteed. This implies that 25%— 25GB of total memory capacity—remains available for bursting if sufficient infrastructure resources are available at the time of need. The provider can use the combination of these settings to throttle back overcommitment from aggressive levels (for example, resource guarantees set to <100%) to more conservative levels (for example, resource guarantees always set to 100%), depending on SLAs in place or fluctuating service loads.