Performance evaluation of checkpoint rollback recovery algorithms in distributed systems
Manzo, William Anthony
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/22893
Description
Title
Performance evaluation of checkpoint rollback recovery algorithms in distributed systems
Author(s)
Manzo, William Anthony
Issue Date
1991
Doctoral Committee Chair(s)
Belford, Geneva G.
Department of Study
Computer Science
Discipline
Computer Science
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
Computer Science
Language
eng
Abstract
Performance evaluation of checkpoint rollback recovery strategies for distributed systems is a field which has not been studied much. Considerable work has been completed in the performance analysis of checkpoint strategies in centralized systems. The necessity for such a study is clear considering the fact that although most distributed algorithms of this kind receive an analysis when presented in the literature, not one of these algorithms is rigorously compared to any other one in a controlled environment. We have chosen four algorithms for our study, each with different approaches to checkpoint placement, storage, synchronization and rollback checkpoint selection synchronization. Both analytic and simulation approaches were taken to carry out the performance study. An analytic model for checkpoint rollback recovery performance analysis in a distributed system is presented here. A basic model for a single node in a distributed system was developed and then a model for a distributed system was constructed using the basic model of the single node to form the components of the network. The basic model relies heavily on previous models for centralized systems. In this study, performance was analyzed using the Weibull distribution which has been shown to model experimental failure data better than the exponential distribution. Finally, the distributed system model was used as a basis for extended general network models that cover the different assumptions of the algorithms studied.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
