Adaptive load migration systems for PVM

Author

Casas, Jeremy ; Konuru, Ravi ; Otto, Steve W. ; Prouty, Robert ; Walpole, Jonathan

Author_Institution

Dept. of Comput. Sci. & Eng., Oregon Graduate Inst. of Sci. & Technol., Portland, OR, USA

fYear

1994

fDate

14-18 Nov 1994

Firstpage

390

Lastpage

399

Abstract

Adaptive load distribution is necessary for parallel applications to co-exist effectively with other jobs in a network of shared, heterogeneous workstations. We present three methods that provide such support for PVM applications. Two of these methods, MPVM (migratable PVM) and UPVM (user-level PVM), adapt to changes in the workstation environment by transparently migrating the virtual processors (VPs) of the parallel application. A VP in MPVM is a Unix process, while UPVM defines lightweight process-like VPs. The third method, ADM (adaptive data movement), is a programming methodology for writing programs that perform adaptive load distribution through data movement. These methods are discussed and compared in terms of effectiveness, usability and performance

Keywords

Unix; message passing; parallel algorithms; programming; resource allocation; virtual machines; ADM; MPVM; UPVM; Unix process; adaptive data movement; adaptive load distribution; adaptive load migration systems; effectiveness; message passing system; migratable PVM; parallel applications; performance; programming methodology; shared heterogeneous workstation network; transparent migration; usability; user-level PVM; virtual processors; workstation environment changes; Adaptive systems; Application software; Computer networks; Computer science; Concurrent computing; Distributed computing; Programming profession; Usability; Workstations; Writing;

fLanguage

English

Publisher

ieee

Conference_Titel

Supercomputing '94., Proceedings

Conference_Location

Washington, DC

Print_ISBN

0-8186-6605-6

Type

conf

DOI

10.1109/SUPERC.1994.344302

Filename

344302