Downloads & Free Reading Options - Results

Includes bibliographical references and index

Includes bibliographical references and index

Characteristics of the resource structure found in the operating system are considered as a mechanism for classifying distributed computer systems. Since the operating system resources, themselves, are too diversified to provide a consistent classification, the structure upon which resources are built and shared are examined. The location and control character of this indivisibility provides the taxonomy for separating uniprocessors, computer networks, network computers (fully distributed processing systems or decentralized computers) and algorithm and/or data control multiprocessors. The taxonomy is important because it divides machines into a classification that is relevant or important to the client and not the hardware architect. It also defines the character of the kernel O/S structure needed for future computer systems. What constitutes an operating system for a fully distributed processor is discussed in detail.

Project EOS is studying the problems of building adaptable real-time embedded operating systems for the scientific missions of NASA. Choices (A Class Hierarchical Open Interface for Custom Embedded Systems) is an operating system designed and built by Project EOS to address the following specific issues: the software architecture for adaptable embedded parallel operating systems, the achievement of high-performance and real-time operation, the simplification of interprocess communications, the isolation of operating system mechanisms from one another, and the separation of mechanisms from policy decisions. Choices is written in C++ and runs on a ten processor Encore Multimax. The system is intended for use in constructing specialized computer applications and research on advanced operating system features including fault tolerance and parallelism.

The applicability of objects architecture for designing real-time reactive operating systems is examined through a model of its elements, their relationships and operations. The requirements of creation, deletion and manipulation of objects in a distributed system with a high degree of fault tolerance are introduced and analyzed. Then, the imposition of time constraints is introduced, and its influence on the objects architecture is analyzed. Relationships with other objects and additional fault tolerance relations are considered. Finally, examples of some solutions to fundamental real-time problems are proposed. Keywords: Systems engineering; Embedding; Computer programming.

Thesis (M.S. in Comp. Sci.)--Naval Postgraduate School, 1978

Thesis (M.S. in Comp. Sci.)--Naval Postgraduate School, 1978

The Star-O/S distributed operating environment was developed to provide communication and synchronization between multiple cooperating expert systems. Each expert system executes as a distinct logical process within the Star-O/S environment with access to multiple Common Lisp environments coordinated by the system.

This report presents the results of the Distributed Operating System Interoperability Research project sponsored by the Air Force Rome Air Development Center (RADC) from June through September of 1987. The goal was to identify guidelines for developing a future integrated Distributed Operating System (DOS) to support a large Battle Management/Command, Control and Communication (BM/C3) system. The project participants drew upon their experience with existing DOSs to determine the key issues involved in designing a future DOS. After comparing BM/C3 requirements and the status of current DOS research, we concluded that no current DOS designs adequately address BM/C3 requirements and additional research should be focused on meeting BM/C3 needs. In particular, significant research should be directed toward evolving a DOS design that can simultaneously provide: Evolvability, Fault Tolerance, Multi-Level Security and Real-Time Processing Support.

This technical report is an attempt to survey the organization principles for distributed systems in a systematic and concise manner. Starting with a comprehensive set of terms covering the area of distributed computing, a classification scheme for distributed operating systems is developed. Based on this classification scheme several communication models are surveyed. Client- server models as an attractive structuring means for distributed operating systems are discussed in greater depth. The report concludes by elaborating the nature of cooperation as an unique underlying principle to organize the work in distributed systems. Distributed computing -as opposed to centralized computing- appears as an attractive alternative to configure a computing system and to organize its work. It receives its potential power from a characteristics fundamental to the structure of any distributed system: its conceptually unlimited extensibility in terms of processing nodes and storage capacity. This property gives raise to the hope that some day one will be able to built distributed systems which -in terms of processing power and reliability- will exceed by far the most powerful centralized systems ever built.

The EOS project is investigating the design and construction of a family of real-time distributed embedded operating systems for reliable, distributed aerospace applications. Using the real-time programming techniques developed in co-operation with NASA in earlier research, the project staff is building a kernel for a multiple processor networked system. The first six months of the grant included a study of scheduling in an object-oriented system, the design philosophy of the kernel, and the architectural overview of the operating system. In this report, the operating system and kernel concepts are described. An environment for the experiments has been built and several of the key concepts of the system have been prototyped. The kernel and operating system is intended to support future experimental studies in multiprocessing, load-balancing, routing, software fault-tolerance, distributed data base design, and real-time processing.

This report looks at the current and the future functionality of the distributed operating systems needed to support Naval combat Systems.

This report looks at the current and the future functionality of the distributed operating systems needed to support Naval combat Systems.

This report looks at the current and the future functionality of the distributed operating systems needed to support Naval combat Systems.

This report looks at the current and the future functionality of the distributed operating systems needed to support Naval combat Systems.

This report looks at the current and the future functionality of the distributed operating systems needed to support Naval combat Systems.

24 p. 28 cm

BBN has developed a operating system, Cronus functions in the context of a heterogeneous internetwork system architecture. Cronus is intended to introduce coherence and uniformity to a set of otherwise independent and disjoint computer systems. Cronus also provides a development methodology, tools and mechanisms to support distributed application development and support for functional properties such as scalability, resource management, and survivability.

The Next Generation Computer Resources (NGCR) Program is the Navy 6.3 program being proposed to provide computer resources to the Navy through the adoption of an open systems architecture. With the initiation of an open system architecture, it is expected that (more) rapid technology transition will occur. The Distributed Operating and Run-Time Systems task of the Mission Critical Computing Project is providing 6.2 research results on which to base the NGCR operating Systems Standards Working Group (OSSWG). The NGCR Program will provide the majority of computer resources to the Navy for military critical applications. Influencing the NGCR is an important payoff for Navy research. By providing a 6.2 research base, this task puts the NGCR OSSWG in a position to make technically knowledgeable decisions in a timely manner. The objective of this task is to address research and technical issues (as identified in the NGCR OSSWG) in the area of distributed operating and run-time systems. Candidate systems will be evaluated with respect to their current level of maturity and to how well they address the NGCR's requirements for an operating system. Section 2 describes the NGCR OSSWG abstract model used as a framework for evaluating the candidate systems. Sections 3 and 4 present Mach and ARTX in this framework. The results are summarized in section 5.

The microprocessor revolution has produced a capable computer on a single printed circuit board. The design and development of a real-time operating system for a distributed system of Single Board Computers is presented in this paper. There are user manuals and program descriptions for the operating system, a debug module, a CRT module and a line printer module. The operating systems has been developed for a Multibus system with three INTEL Single Board Computers SBC80/20-4 and 64K bytes of common memory. The system has been designed specifically for Naval Tactical Data Systems applications and the feasibility of such applications are evaluated with respect to currently available Single Board Computers and with respect to Single Board Computers that should be available in the near future. (Author)

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

Contained in this report are three major design specifications for the ArchOS decentralized operating system that is being built at CMU as part of the Archons project with partial support from NOSC. Each design specification appears as a separate chapter in Section II of this report. The Client Interface Specification in Chapter 4 of this report describes the functionality of ArchOS as seen by client application programs. In Chapter 5, the ArchOS System Architecture Specification summarizes the structure of ArchOS that supports the ArchOS client interface. Chapter 6 contains the System Design Description of ArchOS. The System Design describes the structure and functionality of major components of ArchOS, such as the ArchOS kernel, the arobject/process management subsystem, the transaction subsystem, the file subsystem, the I/O device subsystem, the time-driven scheduler subsystem, and the monitoring and debugging subsystems.

This is the final status report on the DoD-URIP grant for instrumentation to support research in distributed operating systems, fault-tolerant distributed computing, object-oriented computing and other related facets of distributed computing. The Distributed Systems Laboratory was established in the Computer Science Department of the University of Minnesota for this purpose. Currently this laboratory is supporting experimental research in the area fault-tolerant distributed operating systems, specifically to support development of the NEXUS distributed operating system. Currently this laboratory has a total of 12 workstations and 2 file servers with total storage capacity of 2 gigabytes. All these workstations are connected by an ethernet local area network that supports the Computer Science Department.

This is the final status report on the DoD-URIP grant for instrumentation to support research in distributed operating systems, fault-tolerant distributed computing, object-oriented computing and other related facets of distributed computing. The Distributed Systems Laboratory was established in the Computer Science Department of the University of Minnesota for this purpose. Currently this laboratory is supporting experimental research in the area fault-tolerant distributed operating systems, specifically to support development of the NEXUS distributed operating system. Currently this laboratory has a total of 12 workstations and 2 file servers with total storage capacity of 2 gigabytes. All these workstations are connected by an ethernet local area network that supports the Computer Science Department.

This is a working draft of a future reference bibliography on distributed systems, their architecture and principally their controlling software known as an executive or operating system. The general operating systems literature pertaining to those issues which are considered most relevant for distributed systems (i.e., Communications and Synchronization, Protection, Fault-Tolerance, System Structuring Concepts, Programming and Control Languages) has also been included. (Author)

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file. These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

The project 'Portable Operating Systems for Network Computers' that ran from April 1982 until August 1985 saw the creation of an entire distributed operating system (MICROS/SAM2S) for a multiple ethernet system of DEC LSI-11 and Motorola 68000 systems. The modularity, hidden type managers, abstract manager hierarchies, message-oriented drivers, and layered tasks used in the design of the SAM2S operating system allowed the easy porting of the original LSI-11 based system to Motorola 68000 processors and its extension to a plexus of four interconnected thernets. The system was used to explore ways to support efficient communications within groups of processes scattered over many computers, especially in very large networks. This project has carefully delineated different classes of multicast communications within networks and has shown three different efficient waqys to implement multicast: host, channel, and tree-based protocols. Only tree multicast is suitably efficient for networks of thousands of computers.

The project 'Portable Operating Systems for Network Computers' that ran from April 1982 until August 1985 saw the creation of an entire distributed operating system (MICROS/SAM2S) for a multiple ethernet system of DEC LSI-11 and Motorola 68000 systems. The modularity, hidden type managers, abstract manager hierarchies, message-oriented drivers, and layered tasks used in the design of the SAM2S operating system allowed the easy porting of the original LSI-11 based system to Motorola 68000 processors and its extension to a plexus of four interconnected thernets. The system was used to explore ways to support efficient communications within groups of processes scattered over many computers, especially in very large networks. This project has carefully delineated different classes of multicast communications within networks and has shown three different efficient waqys to implement multicast: host, channel, and tree-based protocols. Only tree multicast is suitably efficient for networks of thousands of computers.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.

Password-based user identification and authentication in a network-based operating system generally relies upon a single file that contains user information and the encoded or hashed representations of each users' password. Operating system designers have resorted to various protection schemes to prevent unauthorized access to this single file, These techniques have proved vulnerable to various attacks, the result being unauthorized access to the targeted computer system. This paper proposes a model for a distributed password system in a network environment that eliminates the single password file as a target without introducing additional computational complexity or incorporating additional cost to the user with such items as tokens or biometrics. This application incorporates proven encryption techniques and a distributed architecture to enhance the reliability of an operating system's identification and authentication procedures. The paper provides an object-oriented model of this approach, along with an analysis of a possible implementation in a current operating system.