Seminar by Olivia Das
Performance and Dependability Modeling of Layered Systems
Olivia Das
Ryerson University, Toronto
Date: Monday, April 30th, 2012
Time: 12NOON
Venue: CS102.Abstract:
Most distributed systems with clients and servers are constructed with a layered architecture. In such layered systems, failure of a processor or failure of a process in one layer can cause many other processes that depend on its services to fail, unless they can detect the failure and reconfigure to use a redundant process. Since failures of such systems may incur high cost or may result in loss of life, there has been a great emphasis on designing these systems to be fault-tolerant. In order to achieve the correct and timely operation of such systems, one ought to model them in such a way that they meet both their performance and dependability criteria at the same time, known as performability.
In this talk, I will present a performability model, known as Dependable-LQN model, and its extensions, that can be used for evaluating layered systems. The model can be solved to compute the steady-state and instantaneous measures. The model solution exploits the idea of aggregating the states of a system into multiple classes so that all the states in a class have the same set of effective components (i.e. the set of components which are working and are in use). The model solution avoids the generation and solution of large Markovian models that are commonly used for dependability evaluation.