Brad Glasbergen, Michael Abebe, Khuzaima Daudjee, Daniel Vogel and Jian Zhao receive Best Demo Award at 2020 ACM SIGMOD Conference

Tuesday, June 23, 2020

Cheriton School of Computer Science PhD students Brad Glasbergen and Michael Abebe, along with Professors Khuzaima Daudjee, Daniel Vogel and Jian Zhao, have received the Best Demonstration Award at the 2020 ACM Special Interest Group on Management of Data (SIGMOD) Conference. This year the annual international forum was held virtually from June 14 to 19.

The team of systems and human-computer interaction researchers were conferred the prestigious award for their demonstration of Sentinel — a tool to understand data system behaviour and performance.

photo of Brad Glasbergen, Michael Abebe, Khuzaima Daudjee, Daniel Vogel, Jian Zhao

L to R: Cheriton School of Computer Science PhD students Brad Glasbergen and Michael Abebe, and Professors Khuzaima Daudjee, Daniel Vogel and Jian Zhao

The complexity of modern data systems and applications greatly increases the challenge of understanding system behaviour and diagnosing performance problems. When these problems arise, system administrators have the difficult task of remedying them by relying on large debug log files, vast numbers of metrics, and system-specific tooling.

Sentinel allows administrators to analyze systems and applications by building models of system execution and comparing them to derive key differences in behaviour. The resulting analyses are then presented as intuitive system reports for administrators and developers. Users of Sentinel can find, identify and take steps to resolve the reported performance issues. Sentinel’s models are constructed online by intercepting debug logging library calls, so its functionality incurs little overhead and it works with all systems that use standard debug logging libraries.

Watch “Sentinel: Understanding Data Systems,” a video describing the tool the award-winning research team developed to understand data system behaviour and performance.

The team of ACM SIGMOD judges conferred the Best Demo Award to the Sentinel system for its great contribution to and well-performed demonstration of a not-so-easy-to-show aspect of database systems research — namely the interpretation and comparison of execution logs. Sentinel relies on an abstraction of model behaviour into log events and transitions between events on a per-thread level during program execution. Model behaviours are stored in a database and exported through user and application programming interfaces. The demonstration nicely oscillates between showcasing two scenarios — PostgreSQL and TPC-W workload — and describing the system architecture.

“This is a great achievement for Brad and Michael, my doctoral students,” said Professor Khuzaima Daudjee, a Cheriton School of Computer Science faculty member who works with the Systems and Networking and Data Systems research groups. “Sentinel will be valuable to both system administrators and developers as it simplifies the difficult task of capturing and highlighting important and useful differences in system behaviour. Sentinel’s approach of universally identifying behavioural differences online without performance degradation is a significant step forward, pushing the state-of-the-art for analysis tools. We envision our Sentinel system to pave the way for future research by allowing its users to understand system issues and respond to behaviour changes without administrator intervention.”

The annual ACM SIGMOD/PODS Conference is a leading international forum for database researchers, practitioners, developers and users to explore innovative ideas and results, and to exchange techniques, tools and experiences in all aspects of data management.

  1. 2020 (58)
    1. October (2)
    2. September (5)
    3. August (4)
    4. July (7)
    5. June (4)
    6. May (11)
    7. April (13)
    8. March (3)
    9. February (3)
    10. January (6)
  2. 2019 (90)
    1. December (3)
    2. November (8)
    3. October (6)
    4. September (9)
    5. August (7)
    6. July (8)
    7. June (10)
    8. May (5)
    9. April (11)
    10. March (6)
    11. February (10)
    12. January (7)
  3. 2018 (82)
    1. December (6)
    2. November (8)
    3. October (8)
    4. September (12)
    5. August (4)
    6. July (8)
    7. June (8)
    8. May (5)
    9. April (10)
    10. March (6)
    11. February (2)
    12. January (5)
  4. 2017 (50)
  5. 2016 (27)
  6. 2015 (41)
  7. 2014 (32)
  8. 2013 (46)
  9. 2012 (17)
  10. 2011 (20)