Faculty Recruitment, Mona Attariyan, "Finding the needle in the haystack: How dynamic information flow analysis can help diagnose software misconfigurations"

Contact Name: 
Katie Dahm
GDC Auditorium
Apr 8, 2014 11:00am - 12:00pm
Emmett Witchel

Signup Schedule: http://apps.cs.utexas.edu/talkschedules/cgi/list_events.cgi

Talk Audience: UTCS Faculty, Grads, Undergrads, Other Interested Parties

Host: Emmett Witchel

Talk Abstract: Complex software systems are difficult to configure. When misconfigurations inevitably arise, users spend a considerable amount of time troubleshooting those problems. Misconfigurations are also costly. They cause million-dollar service outages and add substantially to the cost of software support. My work minimizes the impact of misconfigurations by reducing the time to recover from such problems.

In this talk, I will show that system support for dynamic information flow analysis can substantially simplify the diagnosis process by reducing the human effort needed to find misconfigured options. I will present ConfAid and X-ray, two diagnosis tools that automatically pinpoint the likely root causes of a configuration problem. ConfAid focuses on misconfigurations that lead to crashes and incorrect outputs, and X-ray focuses on diagnosing performance misconfigurations. The analysis takes no more than a few minutes, which is much faster and far less labor-intensive compared to manual troubleshooting.

Speaker Bio: Mona Attariyan received her Ph.D. in Computer Science and Engineering from the University of Michigan in 2012, under supervision of Professor Jason Flinn. She joined Google in August 2012 and is currently a Senior Software Engineer working on cloud infrastructure at the Seattle office. She has received multiple awards during her career. She was awarded the University of Michigan Margaret Ayers Host Award in 2010, the Jay Lepreau Best Student Paper Award at OSDI 2012, and the Rising Star in Systems Research Award at the Diversity Workshop at SOSP 2013. She has also received the inaugural ACM SIGOPS Dennis M. Ritchie Best Thesis Award at SOSP 2013.

Her research interests lie in the general area of software systems, with an emphasis on operating systems and distributed systems. Her current research focuses on creating highly distributed and efficient platforms for analyzing life science workloads.