An Efficient Policy-Based Scheduling and Allocation of Virtual Machines in Cloud Computing Environment

Table 2

Overview of the parameters considered for conducted experiments.

Parameters	Test cases
Workload	(i) PlanetLab
	(ii) Google cluster
CPU load of VMs	(i) Constant 10% of the requested capacity
	(ii) Constant 50% of the requested capacity
	(iii) Constant 100% of the requested capacity
VM size	(i) Smaller than in the baseline configuration
	(ii) According to the baseline configuration
	(iii) Bigger than in the baseline configuration
PM power characteristics	(i) All PMs have the same power characteristics
	(ii) Multiple different power characteristics
	(iii) Power characteristics with a slight slope
	(iv) Power characteristics with a large slope
Host overload detection	(i) Static threshold
	(ii) Local regression