Yigong Hu

Postdoc, University of Washington

About Me

I am a postdoctoral researcher working with Prof. Baris Kasikci in the UW Systems Lab at the University of Washington. I earned my computer science Ph.D. from Johns Hopkins University in May 2023 where I worked with Prof. Ryan Huang.

My research is centered around the detection, diagnosis, and mitigation of systems performance issues. In my Ph.D. work, I developed principled approaches to improve the systems performance. Specifically, I contributed to the following areas:

I am currently focused on mitigating intra-application performance interference, employing a range of strategies from developing new scheduling mechanisms to utilizing large language models for assisting in performance issue diagnosis.

Update: I will be an Assistant Professor at ECE@Boston University, starting in Fall 2025. Our lab has multiple Ph.D. student positions. If you are interested in building/hacking systems, we should talk.

News

  • [11/2023] My proposal on diagnosing intra-application performance interference is awarded Azure Cloud Credits.
  • [07/2023] pBox, our work on pushing performance isolation boundaries into application, is accepted to appear at SOSP 2023.
  • [06/2023] I passed my Ph.D. defense and will join University of Washington as a postdoc.
  • [01/2023] vProf(value-assited cost profiling) is accepted to appear at EuroSys 2023
  • [12/2020] Gongqi Huang, an undergraduate student mentored by me, received the honorable mention CRA Outstanding Undergraduate Researcher Award!
  • [08/2020] Our Violet (detecting specious configuration with symbolic execution) is accepted to appear at OSDI 2020.
  • [04/2019] LeaseOS won the best paper award at ASPLOS’19!
  • [11/2018] Our paper LeaseOS, a lease-based resource management mechanism, is accepted to appear at ASPLOS 2019.

Publications

Pushing Performance Isolation Boundaries into Application with pBox
Yigong Hu, Gongqi Huang, Peng Huang
SOSP 2023 [ talk slides software]

Effective Performance Issue Diagnosis with Value-Assisted Cost Profiling
Lingmei Weng, Yigong Hu, Peng Huang, Jason Nieh, Junfeng Yang
Eurosys 2023 [ slides ]

Automated Reasoning and Detection of Specious Configuration in Large Systems with Symbolic Execution
Yigong Hu, Gongqi Huang, Peng Huang
OSDI 2020 [ talk slides software]

Scaling Performance Issue Detection and Diagnosis in Cloud Infrastructures
Yigong Hu, Ze Li, Peng Huang, Suhas Pinnamaneni, Francis David, Yingnong Dang
AAAI workshop on cloud intelligence 2020

A Case for Lease-Based, Utilitarian Resource Management on Mobile Devices [Best Paper Award]
Yigong Hu, Suyi Liu, Peng Huang
ASPLOS 2019 [ talk slides software]

Teaching

[Spring 2022] Guest lecture on early detection of configuration errors for the graduate course, CS 624: Reliable Software Systems ,Johns Hopkins University

[Spring 2021] Guest lecture on detecting misconfiguration with symbolic configuration for the graduate course, CS 624: Reliable Software Systems ,Johns Hopkins University

[Spring 2021] Teaching Assistant for the graduate course, CS 624: Reliable Software Systems ,Johns Hopkins University

[Fall 2019] Invited guest lecture on energy aware android programming for the undergraduate course, EN.601.290: User Interfaces and Mobile Applications

[Fall 2019] Teaching Assistant for the undergraduate course, CS 318/418/618: Principles of Operating Systems

Service

  • Artifact Evaluation Committee Member for EuroSys’23
  • Student Volunteer for EuroSys’23
  • External Review Committee Member for NSDI’21, OSDI’20, SOSP’19, HotOS’19, SOSP’17

Bio

I received my Ph.D. from Johns Hopkins University in June 2023, advised by Prof. Ryan (Peng) Huang. I obtained my B.E. degree in Computer Science from Huazhong University of Science & Technology in 2017. When I was in HUST, I was fortunately worked with Prof. Hai Jin and Prof. Deqing Zou in the Cluster and Grid Computing Lab and Prof. Shan Lu at University of Chicago.

  • My favorite proverb is “The scholar may not be without breadth of mind and vigorous endurance.” from The Analects of Confucius.
  • If I don’t do research, I would spend my time in reading history book, caring for my goldfish, playing Go and most importantly, spend time with my wife and kid.