Skip to content

>AUUG–The Organisation for Unix, Linux and Open Source Professionals

Achieving Parallelism easily through Pshell — Architecture and Overview

by Daniel F. Saffioti


Communications and the scheduling of tasks are the two most important issues of parallel programming on clusters.  Over time, various parallel computing programming models such as remote threads, transparent process migration, message passing, distributed shared memory and optimizing parallel compilers have emerged, assisting the programmer develop applications, which can work seamlessly in such environments.  Considerable research has been done to optimize these models, which typically have large communications overheads, resulting in a detrimental effect on performance.  In addition to this, their acceptance has varied by virtue of the fact that each has introduced new problems with reference to portability, scalability and most importantly usability.  Sometimes these problems completely violate the underlying notion of such computing.

To overcome these issues Pshell, which provides transparent scheduling and communication of jobs between disparate hosts, has been developed.  Pshell is the glue for producing high performance parallel applications, which can work securely, and efficiently in heterogeneous environments.  It represents a major shift in traditional parallel programming environments because it is a language using the syntax of the Bourne Shell sh(1).  The Bourne shell process and communications models can easily be extended to parallel computing environments using the concurrent programming model.

This paper will examine the evolution of cluster computing and identify where deficiencies lie in current programming models, providing a justification for simple languages whilst providing the reader with an understanding of the Pshell programming environment.  In addition to this the paper illustrates how such a language can be used to ease the process of writing parallel applications and to overcome some of the limitations inherent in traditional programming models without scarifying performance.

Daniel F. Saffioti
School of Information Technology and Computer Science (SITACS), University of Wollongong

Download complete paper: Achieving Parallelism easily through Pshell — Architecture and Overview (260K PDF)