Mixing malleable and rigid bandwidth requests for optimizing network provisioning

To address the anticipated sporadic terabyte demand generated by high-end time-constrained applications, dynamically reconfigurable optical networks services are envisioned. However, the time and rate granularities of a bandwidth reservation service and those of transfer tasks using the reserved capacity are not necessarily of the same order of magnitude. This may lead to a poor resource utilization and over-provisioning. This paper explores how request aggregation is able to limit this problem. Interactions between a bandwidth reservation service and a data mover service guaranteeing time-constrained data transfer are investigated. We formulate the underlying optimization problem and propose a linear-program-based strategy for bandwidth provisioning when malleable transfer requests are known in advance. Simulations show that the temporal parameters of requests (deadline and patience) are the dominant criteria and that a small malleability can improve performance a lot.