Need for SPEED: Site Performance Efficiency, Evaluation and Decision

Note: Presentation times are in Coordinated Universal Time (UTC).

Wednesday, 13 October, 2021 - 02:3003:00

Kingsum Chow, Alibaba, and Zhihao Chang, Zhejiang University

Abstract: 

When you are tackling many servers in the data center, saving a small percentage of servers would bring significant return. We will describe how we evaluate performance at scale, and also how it is different from optimization on a single system.

The emergence of large-scale software deployments in the data center has led to several challenges: (1) measuring software performance in the data center, and (2) evaluating performance impact of software or hardware changes. We will highlight a couple of problems that may lead to wrong conclusions. We will present a sketch of our solutions.

Kingsum Chow, Alibaba

Kingsum is a principal engineer at Alibaba CTO Line Technology Risk and Efficiency Group. Since receiving Ph.D. in Computer Science and Engineering from the University of Washington in 1996, he has been working on performance, modeling and analysis of software applications. After working at Intel for 20 years, Kingsum joined Alibaba in 2016. Since then, he has been driving software performance optimization at the scale of data center. He has been issued more than 23 patents. He has presented more than 110 technical papers.

Zhihao Chang, Zhejiang University

Zhihao Chang is a PhD student in the College of Computer Science and Technology, Zhejiang University. His research interests include spatial query optimization.

SREcon21 Open Access Sponsored by Indeed

BibTeX
@conference {276719,
author = {Kingsum Chow and Zhihao Chang},
title = {Need for {SPEED}: Site Performance Efficiency, Evaluation and Decision},
year = {2021},
publisher = {USENIX Association},
month = oct,
}

Presentation Video