Delay Bounds and Asymptotics in Cloud Computing Systems
Add to Google Calendar
With the emergence of big-data technologies, cloud computing systems are growing rapidly in size and becoming more and more complex, making it costly to conduct experiments and simulations. Therefore, modeling computing systems and characterizing their performance analytically is more critical than ever in identifying bottlenecks, informing system design, and facilitating provisioning. In this talk, I will illustrate how we use the association properties of random variables and asymptotic analysis to characterize delay performance. We study the delay of jobs that consist of sub-tasks, where the sub-tasks are processed in parallel on different servers in a computing system, and a job is completed only when all of its sub-tasks are completed. While the delay of individual sub-tasks has been extensively studied, job delay has not been well-understood, even though job delay is the most important metric of interest to end users. Prior work on job delay has investigated the so-called "fork-join" model, but tight analysis is known only for a system with two servers. Finding tight characterizations of job delay in a general system has been an open problem. In our work, we consider a variant of the fork-join model, called "limited fork-join," that is more suitable for modern computing systems. We first provide an upper bound on the job delay, and then show that job delay converges to this upper bound as the number of servers in the system becomes large.
Weina Wang is a joint postdoctoral research associate in the Coordinated Science Lab at the University of Illinois at Urbana-Champaign, and in the School of ECEE at Arizona State University, working with Prof. R. Srikant and Prof. Lei Ying. She received her B.E. from Tsinghua University and her Ph.D. from Arizona State University, both in Electrical Engineering. Her research lies in the broad area of applied probability and stochastic systems, with applications in cloud computing, data centers, and privacy-preserving data analytics. Her dissertation received the Dean's Dissertation Award in the Ira A. Fulton Schools of Engineering at Arizona State University in 2016. She received the Kenneth C. Sevcik Outstanding Student Paper Award at ACM SIGMETRICS 2016.