In this paper, we focus on analyzing application scheduling
through VM and its impact on the energy usage of the cloud
infrastructure in a data center. We have modeled the cloud
infrastructure containing clusters of servers. The model of
each cluster organization and work division of the components
inside the clusters is also designed. To schedule VMs within
a cluster and among the clusters within the whole data center,
we have developed two distributed VM scheduling algorithms.
Our proposed intra- and inter-cluster scheduling algorithms
reduce energy consumption by turning off redundant servers
and keeping under-utilized clusters in sleep mode. Preference
in migrating communicating VMs to same cluster ensures
bandwidth utilization and reduces dealy for communication.
The efficiency and the stability of our algorithms are verified
by simulating it inside a data center using CloudSim toolkit