By increasing the ratio, we can reduce the provisioned capacity and achieve close to 80% utilization.
This task cannot be done without increasing the autonomy of the provisioned Cloud resources.
Furthermore, the provisioned database throughput required depends on future database traffic.
Routing is a key component of this architecture, in which it enables the efficient use of the provisioned bandwidth.
We use the asymptotic growth rate of the provisioned capacity as a metric to compare various design alternatives.
This is where the provisioned resources are sufficient to meet demand and do not go beyond what is required.
In fact, while considering the full life-cycle of cloud services/resources, the efficiency in relation to how end users utilize the provisioned services/resources also matters significantly.
We also find that a sufficiently large fiber-to-node ratio is necessary in order to minimize the asymptotic growth in the provisioned capacity, increase capacity utilization and minimize the need for wavelength conversion.
Moreover, previously proposed methods have treated BE class with lowest priority and if the same priority is given to BE classes in LTE, then half of the provisioned services will suffer from resource scarcity.
IntServ follows the signaled QoS model, in which the end hosts signal their QoS needs to the network using Resource Reservation Protocol (RSVP), whereas DiffServ works on the provisioned QoS model, whereby network elements are set up to service multiple classes of traffic with varying QoS requirements.
The completion time of MapReduce jobs is an important performance metric of Hadoop, especially for the use cases of ad-hoc queries where users need to obtain the results as quickly as possible, and for the use cases of public clouds where users are charged according to the amount of time the provisioned resources are used.
