EFFICIENT PROCESS MIGRATION FOR PARALLEL PROCESSING ON NON-DEDICATED NETWORKS OF WORKSTATIONS
Kasidit Chanchio, He Xian-Sun
Abstract
Open-access reader
Kasidit Chanchio, He Xian-Sun
Abstract
Open-access reader
This paper presents the design and preliminary implementation of MpPVM, a software system that supports process migration for PVM application programs in a non-dedicated heterogeneous computing environment. New concepts of migration point as well as migration point analysis and necessary data analysis are introduced. In MpPVM, process migrations occur only at previously inserted migration points. Migration point analysis determines appropriate locations to insert migration points; whereas, necessary data analysis provides a minimum set of variables to be transferred at each migration pint. A new methodology to perform reliable point-to-point data communications in a migration environment is also discussed. Finally, a preliminary implementation of MpPVM and its experimental results are presented, showing the correctness and promising performance of our process migration mechanism in a scalable non-dedicated heterogeneous computing environment. While MpPVM is developed on top of PVM, the process migration methodology introduced in this study is general and can be applied to any distributed software environment.
OpenAlex reports 7 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents the design and preliminary implementation of MpPVM, a software system that supports process migration for PVM application programs in a non-dedicated heterogeneous computing environment. New concepts of migration point as well as migration point analysis and necessary data analysis are introduced. In MpPVM, process migrations occur only at previously inserted migration points. Migration point analysis determines appropriate locations to insert migration points; whereas, necessary data analysis provides a minimum set of variables to be transferred at each migration pint. A new methodology to perform reliable point-to-point data communications in a migration environment is also discussed. Finally, a preliminary implementation of MpPVM and its experimental results are presented, showing the correctness and promising performance of our process migration mechanism in a scalable non-dedicated heterogeneous computing environment. While MpPVM is developed on top of PVM, the process migration methodology introduced in this study is general and can be applied to any distributed software environment.
Key concepts: Process migration, Computer science, Correctness, Data migration, Workstation, Process (computing), Distributed computing, Scalability