News
Faculty Seminar: A Study of Long-Tail Latency in n-Tier Systems:
RPC vs. Asynchronous Invocations
Dr. Qingyang Wang, Assistant Professor, LSU Computer Science
March 17, 2017 at 3:00 pm, 117 EE Conference Room
Abstract:
Long-tail latency or wide response time fluctuation of web-facing applications continues
to be a serious
problem. An Amazon study showed that every 100ms increase in the page load time decreases
sales by
1%. In this talk, we will describe an experimental study of an important class of
long tail latency problems
that are specific to distributed systems: Cross-Tier Queue Overflow (CTQO) due to
a combination of
millibottlenecks (with sub-second duration) and tightly-coupled servers in n-tier
systems (e.g., Apache,
Tomcat, and MySQL) using RPC-style request-response communications. CTQO is a significant
and broad
problem, since the initiating millibottleneck can originate from any system resource
(e.g., CPU, memory,
and network). We will discuss several practical solutions to this problem. For example,
we show that CTQO
can be reduced or avoided by replacing thread-based servers with asynchronous servers
which support
asynchronous inter-tier communication. Our studies show that in the era of cloud computing
where
resource sharing is a common practice and millibottlenecks are unavoidable, we need
to rethink the
traditional thread-based architecture of n-tier systems in order to achieve both high
performance and high
resource efficiency in cloud.
Bio
Dr. Qingyang Wang is an Assistant Professor in the Department of EECS at Louisiana
State University-
Baton Rouge. His research is in distributed systems and cloud computing with a current
focus on
performance and scalability analysis of large-scale web applications (e.g., Amazon.com).
He has led
research projects at LSU on cloud performance measurements, scalable web application
design, and
automated system management in clouds. He has collaborated extensively with scientists
and researchers
from industry companies, including IBM, HP, Amazon, Fujitsu (Japan), and Wipro (India).
Dr. Wang received
his Ph.D. degree from Georgia Institute of Technology in 2014. He is a recipient of
the Best Student Paper
award in IEEE Cloud 2011 and nominated for the Best Paper Award in IEEE Cloud 2016.