TY - JOUR
T1 - Burstable Instances for Clouds
T2 - Performance Modeling, Equilibrium Analysis, and Revenue Maximization
AU - Jiang, Yuxuan
AU - Shahrad, Mohammad
AU - Wentzlaff, David
AU - Tsang, Danny H.K.
AU - Joe-Wong, Carlee
N1 - Funding Information:
Manuscript received June 22, 2019; revised April 12, 2020; accepted July 13, 2020; approved by IEEE/ACM TRANSACTIONS ON NETWORKING Editor R. La. Date of publication August 26, 2020; date of current version December 16, 2020. This work was supported in part by the NSF under Grant CNS-1751075 and Grant CCF-1453112 and in part by the Air Force Research Laboratory (AFRL) and the Defense Advanced Research Projects Agency (DARPA) under Agreement FA8650-18-2-7846, Agreement FA8650-18-2-7852, and Agreement FA8650-18-2-7862. A preliminary version of this work was presented at the IEEE International Conference on Computer Communications (INFOCOM), Paris, France, 2019. (Corresponding author: Yuxuan Jiang.) Yuxuan Jiang and Danny H. K. Tsang are with the Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Hong Kong (e-mail: yjiangad@connect.ust.hk; eetsang@ece.ust.hk).
Publisher Copyright:
© 1993-2012 IEEE.
PY - 2020/12
Y1 - 2020/12
N2 - Leading cloud providers recently introduced a new instance type named burstable instances to better match the time-varying workloads of tenants and further reduce their costs. In the research community, however, little has been done to understand burstable instances from a theoretical perspective. This paper presents the first unified framework to model, analyze, and optimize the operation of burstable instances. Specifically, we model the resource provisioning of burstable instances, identify key performance metrics, and derive the analytical performance given the resource provisioning decisions. We then characterize the equilibrium behind tenants' responses to the prices offered for different burstable instance service classes, taking into account the impact of tenants' actions on the performance achieved by each service class. In addition, we investigate how a cloud provider can leverage knowledge of this equilibrium to find the prices that maximize its total revenue. Finally, we validate our framework on real traces and demonstrate its usage to price burstable offerings in a public cloud.
AB - Leading cloud providers recently introduced a new instance type named burstable instances to better match the time-varying workloads of tenants and further reduce their costs. In the research community, however, little has been done to understand burstable instances from a theoretical perspective. This paper presents the first unified framework to model, analyze, and optimize the operation of burstable instances. Specifically, we model the resource provisioning of burstable instances, identify key performance metrics, and derive the analytical performance given the resource provisioning decisions. We then characterize the equilibrium behind tenants' responses to the prices offered for different burstable instance service classes, taking into account the impact of tenants' actions on the performance achieved by each service class. In addition, we investigate how a cloud provider can leverage knowledge of this equilibrium to find the prices that maximize its total revenue. Finally, we validate our framework on real traces and demonstrate its usage to price burstable offerings in a public cloud.
KW - Cloud
KW - burstable instances
KW - equilibrium
KW - revenue maximization
UR - http://www.scopus.com/inward/record.url?scp=85090449147&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090449147&partnerID=8YFLogxK
U2 - 10.1109/TNET.2020.3015523
DO - 10.1109/TNET.2020.3015523
M3 - Article
AN - SCOPUS:85090449147
SN - 1063-6692
VL - 28
SP - 2489
EP - 2502
JO - IEEE/ACM Transactions on Networking
JF - IEEE/ACM Transactions on Networking
IS - 6
M1 - 9178492
ER -