Optimal Software Distribution and Allocation Strategies for Improved Datacenter Resource Utilization

Abstract

In this paper, we propose software distribution and allocation strategies to improve the resource utilization for distributed military command centers. For the software distribution strategy, software are divided into multiple slices and simultaneously transmitted along different paths in parallel, with other command centers as cache nodes for shared and collaborated data transit. For the software allocation strategy, we adopt an adapted integer multi-knapsack model to allocate software groups on different servers to utilize the server resource in a load-balanced way for enhanced system stability. The simulation results demonstrate that our distribution strategy achieves 90% time reduction, and software allocation strategy increases 25% resource utilization.