Title: Automatic Undo for Cloud Management via AI Planning
Speaker: Hiroshi Wada, NICTA
Date & Time: Wednesday 17 October 2012, 12:00pm
Location: 310 Soda Hall
Cloud computing provides infrastructure programmatically managed through
simple APIs. This improves the efficiency of system operations; but
having simple powerful system operations may increase the chances of
human-induced faults, which play a large role in system dependability.
To improve the dependability in cloud, it would be helpful if the
platform allows users to rollback to recover from failure. An obvious
approach is to execute a sequence of compensating operations in reverse
chronological order. However, on cloud platforms, this is not always
feasible. Moreover, cloud APIs are often error-prone: we have frequently
observed failures on major commercial cloud platforms. The rollback
therefore must handle failures that occur during the undo. To improve
the dependability of cloud-based systems, we use an AI planner to
automate discovering an appropriate sequence of operations to rollback
the system status. Our planner scales well as the number of operations
needed increases. This work was inspired by our experience in developing
tool support for users of cloud platforms (see http://yuruware.com).
This talk is based on the paper presented at HotDep ´12.
Hiroshi Wada is a senior researcher at Software Systems Research Group,
NICTA. He received a Ph.D. in computer science from University
Massachusetts Boston. His research interests include performance
engineering, distributed computing and bio-inspired computing. He is
also a VP research at Yuruware. At Yuruwware he leads IP development and
product design. Before pursuing a Ph.D., he was with a consulting
company and also represented the Object Management Group in Japan in
technical standpoint. He had also worked as a software engineer in
telecom and application software industry.