High Performance, High Availability Distributed Processor Systems

Distributed processor systems are often heralded as systems which can provide high performance along with availability. In order to realize such an environment it is necessary to develop algorithms for sharing the system workload. Analytic as well as simulation models are being used to study the behavior of various distributed load sharing algorithms. In particular, an optimal deterministic scheduling algorithm has been designed which assigns tasks to processors for minimal execution time. Also a new class of adaptive decentralized load sharing algorithms have been defined and studied using both queueing models and a general distributed system simulator. These models are being used to examine how distributed processor load sharing algorithms behave as a function of job arrival, job service, processor failure, processor repair and interprocessor communication times.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Electrical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2014

Author dc:creator

• Chou, Timothy Chen Kuang

• Engineering, Electronics and Electrical

Language dc:language

eng

Identifier

(UMI)AAI8127567

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/66250