Cost-Aware Traffic Management Under Demand Uncertainty from a Colocation Data Center User’s Perspective

Burstable billing is widely adopted by colocation data center providers to charge their users for transferring. This paper proposes a cost-aware traffic management approach user under burstable where it charged based on the 95th percentile bandwidth usage. To do this, we first develop tractable mathematical expression calculate usage of user. Then, an optimization problem maximize user’s surplus both deterministic and stochastic predictions demand. We show that resulted problem, while non-convex nature, can be efficiently solved or approximated using convex program. also proposed applied in more general scenario gets services from multiple service providers. Using real-world workload traces, reduce IP transit cost 26 percent increase its total 23 percent, compared current practice allocating on-demand.