Welcome to the Order® Lab at Johns Hopkins University (Order® := Operating, Reliable, Defensible, Efficient, Responsive). The lab is directed by Prof. Ryan Huang.
The computing landscape is undergoing exciting changes with emerging hardware, new applications, and unforeseen human factors. Systems, as the backbone of computation and communication, are facing increasing challenges in their dependability.
The overarching goal of the Order Lab is to advance the principles and practices in building next-generation dependable systems for emerging platforms that can foster harmony among systems, applications, developers and users. In the process of achieving this goal, we strive to train world-class systems researchers.
Our interests lie broadly in operating systems, distributed systems, cloud computing, mobile systems, and Internet of Things (IoT).
We are particularly interested in identifying new kinds of issues in these systems such as gray failure in cloud and disruptive mobile app behavior, as well as revisiting old problems in new context such as misconfiguration. While hunting and fighting with bugs is part of what we do, what is ultimately fascinating is an insight that enables principled approaches to tame the wild, harden the fragile, fix the broken, and transcend system reliability and efficiency.
Our research is characterized by three main themes:
Driven by problem rather than solution: our projects are typically driven by observations about some pain point in real-world systems, such as service outages in production cloud services, user-reported app issues, and performance problem in commercial data warehouse system. We spend significant time in understanding a problem rather than jumping to solutions too quickly.
Employ techniques across areas: the solutions we come up with often intersect different areas including system design, program languages, program analysis, software engineering, and machine learning.
Seek real-world impact: we believe the ultimate mission of research is to make an impact in reality. We collaborate closely with leading industry companies and strive to evaluate our solutions in real settings.