The 8th International Workshop on Big Data and Cloud Performance
(DCPerf 2018)
Monday, April 16, 2018 ● 08:00 – 12:00 ● Room: Iolani 3
Cloud data centers are the backbone infrastructure for tomorrow's information technology. Their advantages are efficient resource provisioning and low operational costs for supporting a wide range of computing needs, be it in business, scientific or mobile/pervasive environments. Because of the rapid growth in user-defined and user-generated applications and content, the range of services provided at data centers will expand tremendously and unpredictably. Particularly, big data applications and services, e.g., social and environmental sensing, and IoT monitoring, present a unique class of challenges in the Cloud. In addition, the high volume of mixed workloads and the diversity of services offered render the performance optimization of data centers even more challenging. Moreover, important optimization criteria, such as scalability, reliability, manageability, power efficiency, area density, and operating costs, are often conflicting. The increasing mobility of users across geographically distributed areas adds another dimension to optimizing big data and cloud applications.
The goal of DCPerf is to promote a community-wide discussion to identify suitable strategies to enable effective and scalable performance optimizations. We are looking for papers that present new techniques, introduce new methodologies, propose new research directions, or discuss strategies for resolving open performance problems for hosting big data analytics in the cloud.
Chairs:
Shaolei Ren (University of California, Riverside, USA)
Juan F. Pérez (Universidad del Rosario, Colombia)
___________________________________________________________________________________________________________________________
08:00–08:10
Opening Remarks
Chairs: Shaolei Ren and Juan F. Pérez
08:10–09:00
Keynote Session
Shaping the Clouds from a QoSE (Quality-Open-Smart-Green) Perspective
Speaker: Fangming Liu (Huazhong University of Science and Technology, P.R. China)
Abstract: Cloud computing and cloud storage upon large-scale datacenters (DCs) and content delivery networks (CDNs), are becoming the fundamental paradigm of multiplexing and managing massive computing, storage, networking and big data resources as utility, which host a wide range of Internet-scale services and applications. In this keynote, we envision a four-dimensional development trend of clouds in terms ofQuality-Open-Smart-Green (QoSE), by not only exploring new design space and open source deployment of emerging cloud systems at the service-level spectrum (such as multiple inter-clouds and hybrid clouds), but also identifying challenges and opportunities of underlying DC resource management at the infrastructure-level spectrum (such as SDN/NFV-enabled DC performance guarantee and data-driven energy efficiency optimization). Concrete case studies and inspiring research results will be illustrated to bridge theory and practice, so as to foster comprehensive brainstorming and cross-disciplinary collaboration for shaping the future clouds.
09:00–10:00
Session 1: Experimental Big Data
Online Metrics Prediction in Monitoring Systems
Matthieu Caneill (University of Grenoble Alpes, France); Noel De Palma (Université de Grenoble - France, France); Ali Ait-Bachir, Bastien Dine and Rachid Mokhtari (Coservit, France); Yagmur Cinar (University of Grenoble Alpes, France)
Empirical Study on Taxi's Mobility Nature in Dense Urban Area
Zhenkun Qiu (University of Science and Techonology of China, P.R. China); Sihai Zhang and Wuyang Zhou (University of Science and Technology of China, P.R. China); Shui Yu (Deakin University, Australia)
Available Bandwidth Estimation in Public Clouds
Phuong Ha and Lisong Xu (University of Nebraska-Lincoln, USA)
10:30–12:00
Session 2: Distributed Transactions and Scheduling in the Cloud
Building Efficient and Available Distributed Transaction with Paxos-based Coding Consensus
Shenglong Li, Quanlu Zhang, Zhi Yang, Hanyu Zhao and Yafei Dai (Peking University, P.R. China)
AQM with Multi-queue for Microburst in Data Center Networks
Wataru Morita, Daisuke Sugahara, Kouji Hirata and Miki Yamamoto (Kansai University, Japan)
Stochastic Non-preemptive Co-flow Scheduling with Time-Indexed Relaxation
Ruijiu Mao, Vaneet Aggarwal and Mung Chiang (Purdue University, USA)
Trade-off between Fairness and Efficiency in Dominant alpha-fairness Family
Youngmi Jin (KDDI Reserach, Inc, Japan); Michiaki Hayashi (KDDI Research Inc., Japan)