Showing papers in "Ibm Systems Journal in 1987"

A framework for information systems architecture

Abstract: With increasing size and complexity of the implementations of information systems, it is necessary to use some logical construct (or architecture) for defining and controlling the interfaces and the integration of all of the components of the system. This paper defines information systems architecture by creating a descriptive framework from disciplines quite independent of information systems, then by analogy specifies information systems architecture based upon the neutral, objective framework. Also, some preliminary conclusions about the implications of the resultant descriptive framework are drawn. The discussion is limited to architecture and does not include a strategic planning methodology.

Box structured information systems

Abstract: The box structure methodology for information systems development is based on a usage hierarchy of data abstractions, in which each abstraction is defined in three distinct forms, called its black box, its state machine, and its clear box Each of these three box structures defines identical external behavior, but with increasing internal visibility, to provide a hierarchical structure which supports the systems development principles of referential transparency, transaction closure, state migration, and common services This hierarchy of box structures provides, in turn, a basis for orderly management of information systems development by a finite set of analysis and design tasks in a spiral development process The methodology and its use are described

Geometry for construction and display

Abstract: Matric notation is used to develop geometric concepts for computer-controlled graphics. This natural form of geometric expression leads to homogeneous coordinates which form the basis of an algorithm used for geometric construction. In this way, three-dimensional objects and other pictures can be displayed on a graphics console. The paper also briefly discusses the notation and development of functions for the construction of surfaces.

SNA: current requirements and direction

Abstract: Since its announcement in 1974, Systems Network Architecture (SNA) has evolved in terms of its functional content, configurational flexibility, and network management services. This paper briefly traces this progress to the present and examines the more recent advances in greater detail. It then discusses known requirements for enhanced application and transaction services, for additional provisions for very large networks, for continuing adaptation of small-system and transmission media advances, for inclusion of additional management capabilities, and for further integration of network standards—all of which will shape future SNA developments.

Visual interpretation of complex data

Abstract: With increasingly complex digital simulations and computations, larger volumes of output are generated, and users must select a concise method of displaying the output and extracting relevant information. A set of imaging functions and display modes is developed to interpret data effectively for a wide range of applications. The imaging functions are complementary. Each function is useful for a different aspect of data interpretation. The relationships between variables and the global structures within the data are obtained with different display modes such as multiple windows and animation. With this set of complementary imaging functions and display modes, more information is obtained than with prior imaging methods. Also, more complex simulation studies are feasible since the results can now be visualized.

The structure of System/88, a fault-tolerant computer

Abstract: In recent years, there has been a growing requirement for continuous processing capability approaching 24 hours per day, 7 days per week. Industries such as finance, transportation, securities, and telecommunications have continuous-availability requirements that can approach downtimes of not more than three minutes per year. This paper describes configurations of the Stratus/32 continuous processing computer system that are marketed as the IBM System/88 through an agreement with Stratus Computer, Inc. The system achieves its fault tolerance via hardware duplexing coupled with a distributed operating system that allows system resources to be distributed over many separate computers while maintaining a single systems image to the end user. This single systems image may also be extended across a network of multiple systems. The way in which software makes this distribution possible and the way in which system resources are named to allow transparent distribution across the system are described in the paper. Also described are the transaction processing services that are part of the operating system and allow transaction programs to be written to operate effectively over the distributed system, by means of a requester-server structured approach.

A perspective on advanced peer-to-peer networking

Abstract: This paper is intended to familiarize the reader with the many reasons for undertaking the design and implementation of peer networking on small and intermediate business machines such as the IBM System/36 family. Such networking function was recently announced as IBM's Advanced Peer-to-Peer Networking (APPN) on Release 5 of the System/36. This paper sets the stage for a companion paper in this same issue, which discusses the implementation experience and details of the System/36 APPN product. In the present paper, the history of System/36 communication is first reviewed, and it is shown how APPN was a natural evolution from earlier function. Then an extensive study of user requirements that was started in 1982 is summarized. The paper concludes with a brief technical tutorial on the structure of the APPN design.

OSI-SNA interconnections

Abstract: As Open Systems Interconnection (OSI) becomes an international standard, it is gaining support in both industry and government agencies. One of the major applications of OSI is to act as an intermediary between heterogeneous networks. This paper discusses a scheme for interconnecting a Systems Network Architecture (SNA) network with OSI. This scheme is based on a joint study between IBM Japan and Nippon Telegraph and Telephone Corporation conducted during 1984. Fundamental relationships between OSI session and transport layers and SNA Logical Unit type 6.2 are explored. An OSI-SNA gateway structure is examined, and data units, address translation, and exception handling are discussed.

The IBM RT PC ROMP processor and memory management unit architecture

Abstract: The ROMP processor is the microprocessor used in the IBM RT PC. It is a 32-bit processor with an associated memory management unit implemented on two chips. ROMP is derived from the pioneering RISC project, the 801 Minicomputer at IBM Research. This paper describes some of the trade-offs which were made to turn the research project into a product. It gives an introduction to the architecture of ROMP, including the addressing model supported by ROMP's memory management unit. Some of the unique features of the programming model are explained, with high-level language coding examples which show how they can be exploited. ROMP's architecture is extensible, and the fact that almost all programming for the RT PC has been in high-level languages means that the RT PC hardware architecture can be extended as needed to meet future requirements while preserving the investment in existing software.

A perspective on the 801/Reduced Instruction Set Computer

Abstract: From the earliest days of computers until the early 1970s, the trend in computer architecture was toward increasing complexity. This complexity revealed itself through the introduction of new instructions that matched the application areas. Microcode was an implementation technique that greatly facilitated this trend; thus, most computers were implemented using microcode. In 1975, work began at the Thomas J. Watson Research Center on an experimental minicomputer. This project, termed the 801 project, questioned the trend toward complexity in computer architecture. It was observed that most of the complex instructions were seldom used. Thus, a computer could be designed with only simple instructions without drastically increasing the path length or number of instructions required to implement an application. This made it possible to implement a machine without resorting to microcode, which improved performance. This paper described the background and evolution of these ideas in the context of the 801 experimental minicomputer project.

A large-scale computer conferencing system

Abstract: This paper discusses the relationships between computer-mediated communications and other forms of communication and describes a particular computer conferencing system in use within IBM. The system described is quite large, with over three thousand contributors and over twenty thousand readers. We discuss the structure of the system, the actions that users can take, and the ways in which the system is being used. Neither the definitions presented nor the system described are intended to be the last, or only, word on the subject; as computer-mediated communications and distribution become more and more important in the business and professional communities, we will need more ways of thinking about communication systems and about information distribution in general.

Implementing System/36 advanced peer-to-peer networking

Abstract: System/36 Advanced Peer-to-Peer Networking (APPN) provides highly dynamic, fully distributed peer networking for low-end processors. It is built upon existing SNA Logical Unit 6.2 and Node type 2.1 support. APPN presents System/36 users with a simplified model of communications. The structure of the APPN subsystem is outlined, with particular emphasis on the integration of APPN functions with existing SNA support. The authors describe how particular aspects of the APPN design have been tuned to the System/36 operating environment.

Specification and implementation of an ISO session layer

Abstract: This paper describes a novel technique for the specification and implementation of layered communication software. The technique is called Parallel Activity Specification Scheme (PASS) and is based on an extended-state machine model of protocol automata. It allows a convenient description of the communication behavior of concurrent systems and semiautomatic generation of programming language code from the specification. The first large-scale experience gained with this technique was in the specification and implementation of an ISO session layer. The code generation process and the embedding of the session code into a portable OSI operating system environment are described in detail.

Structures for networks of systems

Abstract: This paper describes how systems will be interconnected in the future, the roles that they will play, and the trade-offs that affect these roles. Starting with a general model for structuring a network of systems, general trade-offs in cost and performance are discussed relative to where functions are placed in the network. Several general principles for data and function placement in a network of systems are derived from these trade-offs. The optimal roles for each of several layers of a network of systems are discussed. Finally, conclusions are drawn regarding the design of future networks of systems.

Message-handling systems based on the CCITT X.400 recommendations

Abstract: Message-handling systems allow the exchange of electronic mail between computers. The International Telegraph and Telephone Consultative Committee (CCITT) has proposed a standard for message-handling systems in the form of the X.400 series of recommendations that has been widely recognized by computer manufacturers and communications carriers. This paper provides a tutorial on the X.400 recommendations and then describes two prototypes developed by the IBM European Networking Center in Heidelberg, Germany, in cooperation with its research partners. The prototypes were demonstrated together with X.400 prototypes from other manufacturers at the CeBIT 86 trade fair in Hannover, Germany.

An incidence-matrix-driven panel system for the IBM PC

Abstract: A set of programs called TRYLON is discussed which permits the application developer to design and create a set of intelligent full-screen panels. These panels serve as a user interface for application programs. The panels and the linkages between them are uncoupled from the application code, thereby reducing programming effort and development time. An incidence matrix is used to describe the graph of a network composed of the panels. Because the paths among the panels are specified by entries in the incidence matrix at panel creation time, there is no need to program the logical constructs for the network.

A page-swapping prototype for VM/HPO

Abstract: This paper discusses a series of changes that were made to a system running the Virtual Machine/System Product with the High Performance Option to enhance paging. The motivation and the background for these enhancements are discussed, and the design of a series of experimental paging subsystems is described and contrasted with the old design: specifically, the new algorithms for main memory management, block paging, working set identification, trimming, prepaging, page replacement, page-out device selection, and page-out slot selection. The performance impact of these changes is illustrated by results of benchmark measurements, which are then contrasted to measurements without the enhancements. Some things learned in running the prototype are discussed and conclusions drawn.

Prospect and design choices for integrated private networks

Abstract: This paper reviews some of the choices that will be available in the next few years, as the much-discussed move toward implementing voice and data integration within a single wide-area integrated private network proceeds. After the term :hp1de-area integrated private network:ehp1. has been defined, a discussion is given of requirements the network ought to satisfy for its users. Then two particularly promising approaches, fast packet switching (FPS) and hybrid switching (HS), are defined, and specimen design points for FPS and HS are postulated, so that the two can be compared. While a definitive comparison would require systematic cost and performance studies, much insight can be gained from the qualitative comparison that we present here. We assess some of the arguments that have been put forward in favor of FPS or HS and conclude that, although today both architectures have promise, and research on both should continue, FPS appears to be slightly simpler to implement and operate.

Advanced interactive executive (AIX) operating system overview

Abstract: The Advanced Interactive Executive (AIX) is the operating system used in the RT Personal Computer. It is a portable operating system architecture that is suitable for a wide range of computer architectures and customer requirements. Discussed in this paper are the structure and services of AIX.

Advanced interactive executive program development environment

Abstract: The IBM RT Personal Computer uses the Advanced Interactive Executive as an operating system. This operating system provides a distinct environment for the development of programs. Some of the characteristics of application development with this operating system, some of its features that influence application design, and the basic program development tools are described.

Data communications: the implications of communication systems for protocol design

Abstract: The construction of a communication network architecture, specifying protocols by which systems communicate, is a complex art. Much has been written about the optimal protocols for theoretical models of systems. This paper points out that protocol design must depend on the “nuts and bolts” of the systems which implement the protocols. Numerous examples are provided to support this thesis. The paper also briefly discusses other issues that influence protocol design and draws lessons for standards activities.

AIX usability enhancements and human factors

Abstract: As microcomputers become capable of running increasingly large and complex operating systems, the question of the usability of those operating systems becomes critical. Most microcomputer users neither are nor want to be systems programmers, yet most of the existing large operating systems assume the existence of a dedicated systems programming organization to install and maintain system software. This paper describes the process by which a large existing operating system was modified to allow it to be installed, configured, maintained, and used by individuals with minimal programming knowledge. We describe the aspects that had to be changed, the kinds of modifications that were required, the reasoning behind those modifications, and the priorities that constrained our activity. We also describe the development process by which potential usability problems were identified and corrections were defined, implemented, and validated.