A. Goscinski

A remote process creation and execution facility supporting parallel execution on distributed systems

The parallel execution of processes over a distributed system is an objective, that, if realised, will provide access to a wealth of computational resources currently underutilised in many network environments. What is lacking from current distributed systems is a flexible and user friendly facility for the creation of processes across a network of workstations in a transparent and easily controlled manner. The paper presents an original architecture of a remote process creation and execution facility designed and developed as part of a microkernel and client-server based distributed operating system. The most important feature of this facility is that it coordinates the cooperation of other kernel servers to create (locally or remotely) a process. It was demonstrated that this facility is very fast and, thus can be used efficiently to support parallel execution on a network of workstations.

Performance comparison of process migration with remote process creation mechanisms in RHODOS

We claim that both remote process creation and process migration are efficient mechanisms to be used in the improvement or development of high performance computer systems. In particular we demonstrate that the claims made by some researchers that process migration is too heavy to be used to support dynamic load balancing are unsubstantiated. We support our claim by presenting these two mechanisms available in the RHODOS distributed operating system, comparing and contrasting these mechanisms and reporting on their performance.

Implementation and performance of the interprocess communications facility in RHODOS

The paper examines the implementation and performance of the RHODOS interprocess communications facility. This examination includes; the Local IPC module provided by the microkernel and the primitives and mechanisms that it supports, the Interprocess Communications Manager, (IPCM), and the Network Manager, in particular the remote and local interprocess communications performance. Furthermore we discuss the role of the message buffers in the interprocess communications facility and their effect on communications throughput.

Preliminary comparison between the single-threaded and multi-threaded RHODOS microkernel

RHODOS is a distributed operating system based on the client server model and a microkernel paradigm. It is a research tool designed to be capable of implementing and evaluating different mechanisms and policies that are proposed by researchers. To maintain suitability for this task, the RHODOS microkernel has recently been upgraded from single to multi threaded operation. However, this upgrading has resulted in a significant number of changes to the operation of some system calls and interrupt handlers. The paper compares the two versions of the RHODOS microkernel, and quantifies the changes that have been made with performance timings. This work is very important, as it permits researchers to understand the characteristics and performance limitations of the user and kernel processes they create.

Towards a framework for aligning learning outcomes, academic literacies and assessment criteria

Although Australian students spend three or more years studying they can seem quite unaware of any of the expected learning outcomes of their course. They are often single unit focused, paying most attention to individual assessment items thus not developing a holistic view of their course. This paper presents a theoretical framework to support staff and students to recognise, scaffold and achieve learning outcomes and academic skills at unit level and to recognise how these contribute to course and graduate learning outcomes, within the boundaries of Australian university and professional accreditation requirements. A case study is described that demonstrates the manual implementation of the framework. The complex nature of the implementation suggests that a software solution is required to ease the process and ensure the resulting mapping will have some longevity by being maintainable.

Toward a Comprehensive Software Based Dsm System

Software based Distributed Shared Memory (DSM) systems have been the focus of considerable research effort, primarily in improving performance and consistency protocols. Unfortunately, computer clusters present a number of challenges for any DSM systems that are not solvable through consistency protocols alone. These challenges relate to the ability of DSM systems to adjust to load fluctuations, computers being added/removed from the cluster, to deal with faults, and the ability to use DSM objects larger than the available physical memory. We present here a proposal for the Synergy Distributed Shared Memory System and its integration with the virtual memory, group communication and process migration services of the Genesis Cluster Operating System.

The architecture of and support provided by the RHODOS distributed file facility

We propose an architecture for the distributed file facility. It explores the potentials of the client server model and is fully scalable. We also report on the support provided to redirection of standard I/O, creation of medium weight processes and migration of processes by this facility.