Many emerging scientific and industrial applications require transferring multiple terabytes of data on a daily basis. Examples include pushing scientific datasets from particle accelerators/colliders to laboratories around the world, synchronizing data centers in different continents, and replicating collections of high definition videos from Olympic Games, taking Location on a different time-zone. A convenient property of all above applications is their ability to tolerate delivery delays from a few hours to a few days. Such Delay-Tolerant Bulk (DTB) transfers are currently being serviced through the postal system using hard drives and DVDs, or through expensive dedicated networks.
In this work we develop store-and-forward (SnF) scheduling policies for achieving low-cost transfer of DTB data over existing public networks. We use traffic data from 200+ links of a large transit ISP to show that the naive approach of using end-to-end (E2E) connection oriented transfers can be prohibitively expensive under common percentile charging schemes. The problem is that despite the existence of strong diurnal load patterns that leave lots of free bandwidth during off-peak hours, E2E transfers are unable to provide low end-to-end cost because very often the load “valley” of a sender does not coincide in time with the load valley of an intended receiver. By utilizing network storage at main PoPs, our SnF policies time-shift DTB transfers in order to bridge the gap between non coinciding valleys, and thus reduce, even zero, the incurred transit costs.
Similarly, under a flat-rate pricing scheme, complementarities in the available bandwidth at the two end points of a DTB flow that crosses multiple time zones may limit the end-to-end available rate to a very small value. In this case SnF scheduling can be used to increase the perceived effective transmission rate.
Who is Nikolaos Laoutaris?
Nikolaos Laoutaris is a research scientist at Telefonica Research, Barcelona; prior to that he was a post-doctorate Fellow at Harvard University (USA) and a Marie Curie post-doctorate Fellow at Boston University (USA). He holds a Ph.D. degree in Computer Science from the University of Athens, Greece (2004). His main research interests are on system, algorithmic, and performance evaluation aspects of computer networks and distributed systems with emphasis on content distribution, overlay networks, P2P, and multimedia communications.
The conference will be conducted in English