Network Scheduling for Big Data Applications
Speaker: Hong ZHANG
Department of Computer Science and Engineering
Hong Kong University of Science and Technology
Title: "Network Scheduling for Big Data Applications"
Date; Monday, 12 March 2018
Time: 4:00pm - 5:00pm
Venue: Lecture Theater F (near lifts 25/26), HKUST
We live in an era with unprecedented amounts of data. To facilitate big
data applications which make sense of the collected data, people have
built big data centers with thousands of machines, and developed
distributed data analytics systems such as Hadoop and Spark. Communication
is playing a significant role in this parallel-computing paradigm, and
networking scheduling has embraced new challenges with the emergence of
these big data applications.
My PhD research has centered around how to identify, understand then
address these new challenges in network scheduling, which in turn leads to
significantly better application performance. In this talk, I will talk
about how network scheduling can facilitate big data applications from two
aspects: rate allocation and path selection. More specifically, I will
introduce the new challenges we identified in both aspects, and
demonstrate how we manage to solve them via application-aware scheduling
and uncertainty-resilient load balancing.
Hong Zhang is a PhD student in the CSE department of HKUST, working with
Professor Kai Chen. Hong is broadly interested in problems related to data
center networking, and specializes in network scheduling and optimization.
He is a recipient of the Google PhD Fellowship in Systems and Networking.