Uncertainty in Clouds: Optimization Challenges and Opportunities
In this talk, we discuss the opportunities and challenges of mitigating uncertainty in cloud computing. We analyze the structure of uncertainties arising from performance and bandwidth changing, virtualization, and elasticity, among others. We also describe uncertainty associated with workload properties, dynamism of the execution context, and uncertainty associated with such important aspects as privacy, security, and availability. We address scheduling problems for different scenarios of HPC, Grid and Cloud Infrastructures. We provide some theoretical and experimental results and discuss static, dynamic and adaptive approaches. We discuss challenges of resource optimization in the presence of uncertainty, ranged from handling resource heterogeneity, dynamic behavior of the execution context as well as uncertainties associated with cloud storages and Internet of Things.
Andrei Tchernykh is holding a full professor position in computer science at CICESE Research Center, Ensenada, Baja California, Mexico. He is Chair of the Parallel Computing Laboratory at CICESE, Mexico and “Laboratory of Problem-Oriented Cloud Computing” at South Ural State University, Russia. He is leading researcher of ISP RAS – Institute for System Programming of the Russian Academy of Sciences and MIPT – Moscow Institute of Physics and Technology, Moscow, Russia. He obtained his Ph.D. in 1986 from the Institute of Precision Mechanics and Computer Engineering of the Russian Academy of Sciences, where he took part in the supercomputer ELBRUS design and development. He gained industrial experience as supercomputer design team leader in Advance Technical Products Corp, and Supercomputer Design Department of Electro-Mechanical Enterprise, Russian leaders in HPC design and development. Tchernykh leads a number of research projects and grants in different countries funded by CONACYT, NSF, ANII, Ochoa, INRIA, FNR, UC MEXUS, DAAD, LAFMI, UJF, INPG, REDII, FUMEC, etc. He has published over 200 papers in some of the most distinguished scientific journals and international conferences and served as a TPC member, and general co-chair of more than 250 professional peer-reviewed conferences. He has graduated 36 Ph.D. and M.S. students, and served as the External Examiner of Academic Council for Ph.D. programs in India, Malaysia, Germany, Luxembourg. México, and France. He is awarded Global Scholars Fellow at Tsinghua University (China), German Academic Exchange Service fellowship at University of Göttingen, Dortmund University, Technische Universität Clausthal, and Severo Ochoa fellowship at Barcelona Supercomputing Center (Spain). He was an Invited Visiting Researcher at Centre de recherché INRIA Lille – Nord Europe (France), Université Grenoble Alpes (France), Luxembourg University (Luxembourg), Moscow Institute of Physics and Technology (Russia), Institut National Polytechnique de Grenoble (France), University of California–Irvine (USA), University of Southern California (USA), Université Joseph Fourier (France), UdelaR (Uruguay), etc. He is an editorial board member of several journals, such as International Journal of Metaheuristics, Supercomputing Frontiers and Innovations, Computational Mathematics and Software Engineering, Proceedings of ISP RAS, etc. He also has served as a guest editor for several special issues including International Journal of Approximate Reasoning (Elsevier) Special Issue on Uncertainty in Cloud Computing: Concepts, Challenges, and Current Solutions. Prof. Andrei Tchernykh is engaged in extensive research on grid and cloud research addressing resource optimization, both, theoretical and experimental, cybersecurity, uncertainty, scheduling, multi-objective optimization, heuristics and meta-heuristics, adaptive resource allocation, energy-aware algorithms and Internet of Things (http://usuario.cicese.mx/~chernykh/)