Infrastructure-as-a-Service (IaaS) provides shared computing resources that users can access over the Internet, which has revolutionized the way that computing resources are utilized. Instead of buying and maintaining their own physical servers, users can lease virtualized compute and storage resources from shared cloud servers and pay only for the time that they use the leased resources. Yet as more and more users take advantage of this model, cloud providers face highly dynamic user demands for their resources, making it difficult for them to maintain consistent quality-of-service (QoS). We propose to use price incentives to manage these user demands. We investigate two types of pricing: spot pricing and volume discounts. Spot pricing creates an auction in which users can submit bids for spare cloud resources; however, these resources may be withdrawn at any time if users’ bids are too low and/or resources become unavailable due to other users’ demands. Volume discount pricing incentivizes users to submit longer-term jobs, which provide more stable resource utilization. We provide insights into these pricing schemes by quantifying user demands with different prices, and design optimal pricing and resource utilization strategies for both users and cloud providers.
Speaker Biography
Carlee Joe-Wong is an assistant professor in the ECE department at Carnegie Mellon University, working at CMU’s Silicon Valley Campus. She received her Ph.D. from Princeton University in 2016 and is primarily interested in incentives and resource allocation for computer and information networks. In 2013–2014, Carlee was the Director of Advanced Research at DataMi, a startup she co-founded from her data pricing research. She received the INFORMS ISS Design Science Award in 2014 and the Best Paper Award at IEEE INFOCOM 2012, and was a National Defense Science and Engineering Graduate Fellow (NDSEG) from 2011 to 2013.