Abstract:

There is no free lunch for application performance: the smaller the response time and/or the higher the throughput, the higher the cost. In a commercial cloud computing environment, capacity planning essentially becomes cost planning, provided cloud computing providers deliver what they promise. In this talk I will use examples to illustrate the tradeoff between performance and cost using simple performance models. It shows that with explicit cost structures of cloud computing, capacity planning becomes a budget planning and cost optimization process for IT organizations embracing the technology.

Biography:

Dr. Yiping Ding is a Vice President of R&D at OPNET Technologies, a leading software provider on performance monitoring, analysis, and prediction of applications, networks, and servers. He is responsible for Advanced Analytical Technologies at OPNET. Before joining OPNET, Yiping was a Chief Scientist at Akorri, a company focusing on managing the virtualized data center. He also worked for BGS Systems and BMC Software as a scientist, Sr. software development manager and architect in the areas of performance modeling and capacity planning for distributed and virtualized platforms. He has authored and coauthored over 50 papers in various technical journals and proceedings and is a frequent speaker at international conferences, industry user group meetings, and trade shows. He holds many US patents in the performance analysis and capacity management area.

Yiping graduated from the Chengdu Institute of Radio Engineering (CIRE) in 1982 (now University of Electronic Science and Technology of China) and taught computer science courses at CIRE afterwards. In 1985, he won a government merit scholarship to pursue graduate degrees in the United States. He obtained M.S and Ph.D. degrees in Computer Science from the University of Connecticut in 1987 and 1991, respectively.