Capacity Planning of Enterprise Information System through Simulation

Enterprise information systems (EIS) nowadays experience the problem of over maintenance and increased operational costs due to the installation of many servers to accommodate either new or existing projects/applications. Most of the servers are underutilized and consumed more electrical power due to cooling and operative expenses. In this paper, we have proposed a server consolidation scheme to improve the servers' utilization, whereby the lowest utilized servers in EIS are removed and their clients are rerouted to the remaining servers with acceptable utilization and network delay. We have utilized two random distribution approaches to select the servers for distributing the clients: single server selected at random and multiple servers selected at random. We have validated our proposed consolidation scheme by simulating the EIS using OPNET IT Guru and analyzing the utilization of servers before and after consolidation. The simulation results demonstrate that the multiple distributions show better performance with improved individual server utilization ranging from 27% to 37% and with 16% reduced overall annual operational cost of EIS.