Abstract:

Configuration errors are among the dominant causes of service-wide, catastrophic failures in today's cloud and datacenter systems. Despite the wide adoption of fault-tolerance and recovery techniques, these large-scale software systems still fail to deal effectively with configuration errors. In fact, even tolerance/recovery mechanisms are often misconfigured and thus crippled in reality.

In his talk, Xu will present research efforts for protecting cloud and data center systems against configuration errors. He will also discuss research that seeks to understand the fundamental causes of misconfigurations.

Then, Xu will focus on two of his approaches: PCheck and Spex. PCheck generates checking code to help systems detect configuration errors early, and Spex exposes bad system reactions to configuration errors based on constraints inferred from source code. Together, these approaches enable software systems to anticipate and defend against configuration errors.

Bio:

Tianyin Xu is a Ph.D. candidate in the Computer Science and Engineering department at the University of California, San Diego. His research areas blend systems, software engineering, and human-computer interactions to address the overarching goal of building reliable, secure systems.

His dissertation work impacted the design and implementation of real-world commercial and open-source systems. In 2016, he and his collaborators won a best paper award at the annual USENIX Symposium on Operating Systems Design and Implementation conference.