Skip to main content

Über dieses Buch

The present volume and volume I "Hector: New Ways in Education and Research" present the results of HECTOR, the four year cooperation project between the Univer­ sity of Karlsruhe and IBM Germany (represented by the European Networking Center and Scientific Center in Heidelberg as well as IBM Research in Riischlikon). The pro­ ject was started in spring 1984 and will end in April 1988 with a congress. This congress addresses the scientific community to present experiences and results with a program of lectures and demonstrations. The HECTOR Project has two major aspects: the first is to explore new ways in uni­ versity education. The second aspect of HECTOR comprises basic research work to develop new technological concepts for the establishment of computer communication networks, supporting academic research and education in all disciplines. The underlying concept is that now and in the future, computer, software and commu­ nication systems which are required for the broad range of scientific and educational tasks will be of different technical orientation and made by different manufacturers. These diverse systems will, however, need to coexist and cooperate side by side. Today, in most cases, different hardware and software architectures of different manu­ facturers prevent a scientist or student from choosing freely the computer and software which offers the best alternative for solving his or her current problem. The mutual cooperation of the academic users is also hindered substantially by the many incompa­ tibilities present. The users' future is therefore transparency in a heterogeneous envi­ ronment.



General Overview


HECTOR Project

Objectives, Organisation, Experience
Data Processing at German universities has changed a great deal in recent years. Starting from supercomputers with many megaflops down to personal computers allowing for scientific work at low cost but also for satisfying high requirements in education. In between there are large and medium sized universal computers in computing centers and in many institutes.
Bernd Krause

LAN and OSI Based Heterogeneous Campus Networks

Until two decades ago the computing center (CC) of a complex organisation, such as a university, consisted of a central computer which was administered and operated by a department devoted exclusively to this task. The users developed programs and used data and equipment stored in the CC in close cooperation with the CC to process and present their results. It soon became clear that many procedures and programs, equipment and data — in short resources — would be useful to a larger community of users, and that the productivity of research and education could be improved if these resources were to be collected by the CC and made generally available to the community.
G. Müller, G. Krüger

Distributed Academic Computing (Project F1)



Computing at academic institutions in the early eighties was still determined by large hosts, department computers and computers in laboratories. The installations of local area networks and personal computers or workstations was just at its beginning. Many universities had a pilot project for a local area network and a plan for a campus network. The interconnection of distributed computing systems was no longer a question of economy. Thus connectivity at the physical level was in sight. Also, at about the same time, ISO published the standards necessary for the exchange of data between heterogeneous computers.
O. Drobnik, H. Schmutz, H. Wettstein

Cooperative Processing in Heterogeneous Computer Networks

This paper introduces the Distributed Academic Computing Networking Operating System as a systematic approach to conquer heterogeneity and to achieve owner controlled resource sharing in computer networks. The major design constraints, like preserving of the investment and maintaining the autonomy of node owners, are described. The paper discusses the principles which led to a consistent design. The components of the system and their relationship are presented and the experiences collected during design and development are reported. Subsequent papers in this volume describe the components of the network operating system in more detail.
H. Wettstein, H. Schmutz, O. Drobnik

A Portability Environment for Communication Software

This paper describes a portability environment, which has been developed for a portable network operating system. The environment has to support methods of structuring communication software in multiple threads and has to provide access to lower level communication services in a uniform, guest system independent way. In order to save the users investments in communication-equipment and -software a method is introduced, which integrates multiple networks into a global net. The requirements of a portability environment are analyzed and the developed design concepts are derived. Furthermore alternatives of its implementation in guest systems are discussed. Finally experiences are presented, which have been gained during its portation from the development system to other guest systems.
R. Staroste, H. Schmutz, M. Wasmund, A. Schill, W. Stoll

Generic Support for Distributed Processing in Heterogeneous Networks

The apparent complexity of distributed application development, especially in a heterogeneous environment, is the prime motivation for the network operating system kernel described in this paper. The kernel reduces this complexity by separating the distribution related issues from the application related ones. It provides an interface of generic objects and operations, which are able to take away from the application programmer most of the problems of distribution, access protection, resource management, and data representation. This paper develops the basic concepts of the kernel from rather general design objectives and illustrates its use and major properties. A prototype implementation, which is running on three different architectures, demonstrates the feasibility of adding these facilities to a given operating system without affecting existing interfaces or applications. The paper reports about early experience with the implementation and performance of the prototype.
H. Eberle, K. Geihs, A. Schill, H. Schmutz, B. Schöner

Directories and Orientation in Heterogeneous Networks

Naming and binding of objects must be considered as central functions of an operating system. In the past, name management has developed from simple name space mapping to directory systems supporting a wide variety of system functions. Naming has gained new attention in the context of distributed systems. During the last years, a great number of distributed services have been realized, each one supported by its own associated directory. To avoid the undesired functional redundancy introduced by such service specific directories, universal directory systems are required capable of supporting any current and future service. This paper describes the concept of the universal directory system developed and implemented as part of the DAC Project.
Bruno Mattes, Heiko von Drachenfels

Authentication and Authorization in Resource Sharing Networks

Authentication and authorization turned out to be of central concern in resource sharing networks. It must be expected, that the willingness of an institution to contribute its equipment to such a network greatly depends on the protection mechanisms provided. We start with discussing a number of alternative approaches and solutions for the authorization and authentication problem. The conclusions drawn from these discussions form the basis for the design of the DAC Authentication and Authorization System which is described next. Emphasis has thereby been laid on the clearness of concepts and on operational robustness of the resulting system.
Bruno Mattes

Transparent Access to Remote Files in Heterogeneous Networks

Computer users share information by accessing common files. In computer networks, the sharing of files requires access to remote files. In heterogeneous networks, users are faced with a broad range of different file systems.
This paper describes an approach for the transparent access to remote files in heterogeneous networks. A unified global file system is derived from common properties of different local file systems. Access transparency is achieved with mapping mechanisms between the local and the global file system. These mechanism and their use are explained. Finally, some aspects of the implementation of such a system on three different operating systems are discussed.
U. Hollberg, E. Krämer

A Remote Execution Service in a Heterogeneous Network

The DACNOS Remote Execution Service enables the sharing of resources such as algorithms, special hardware processors or special peripheral devices. It runs in a heterogeneous network with respect to both hardware and operating systems. Its design being guided by the principle that remote program execution should be a natural extension of local program execution, the DACNOS Remote Execution Service is control transparent at the command level as well as rather strongly home environment transparent. This paper describes the key design aspects of remote execution services in general, and the design and implementation of the DACNOS Remote Execution Service in particular.
Rainer Oechsle

Controlling Distributed User Tasks in Heterogeneous Networks

The Task Setup Service (TSS) offers an interface to the user to run his tasks in a heterogeneous network. In this context the notion of ’task’ comprises interactive sessions as well as batchjobs or more complex applications that have already been existing in one of the single systems or will be developed especially to take advantage of the distribution. While invoking TSS commands the user may, but need not know about the locations of programs, resources and systems that are able to execute some or all parts of his task. A special command supports user-controlled migration to achieve better performance and availability. To achieve this controlling of user tasks, the TSS has to care for the cooperation of DAC services. In the following the integration of the TSS into the overall DAC architecture and its implementation are described.
Cora Foerster

OSI Communication Services for Heterogeneous LAN-Based Environment (Project F2)



The aim of OSI standards is to facilitate communication and cooperation in multi-vendor environments. The standardization of communication services and protocols follows the OSI Reference Model, which defines the general concepts of a layered communication architecture consisting of seven functional layers. For several reasons, the description of services and protocols may not contain specifics of implementation or technology. However, this fact renders their application in real systems more difficult and may delay or even prevent their widespread employment.
L. Svobodova, O. Drobnik

OSI Communication Services in a Heterogeneous Environment

An experimental OSI transport service for a heterogeneous distributed environment based on interconnected local-area networks has been developed as a joint project between the University of Karlsruhe and the IBM Zurich Research Laboratory. The objectives of this effort were twofold:
Provide a reliable and efficient communication base for resource sharing and distributed applications, and
develop in-depth understanding of the issues important to implementation and operation of OSI protocols in such an environment.
The transport service facilitates communication among IBM S/370s, IBM PCs, and DEC VAXs. A network-level Interworking Unit has been developed to interconnect the IBM Token Ring, Ethernet, and X.25 subnetworks. In the associated research projects, the performance of the transport service and the impact of internetworking have been studied both experimentally and via simulation. A comprehensive set of test tools has been developed and applied to test the compatibility of the different implementations of the transport service.
O. Drobnik, L. Svobodova

Implementation of the OSI Transport Service in a Heterogeneous Environment

This paper focuses on the OSI transport service and protocols and their implementation in our experimental OSI communication system for LAN-based distributed environments. After an overview of the transport layer service elements and protocol functions, design issues and related decisions are discussed, along with experiences gathered with our two implementations for IBM S/370 hosts and PC workstations. The issues covered include program/process structuring, inter-layer interfaces, and interactions with the local operating system. Finally, we discuss an enhanced datagram service built on top of the OSI transport service, used for a distributed computing project.
Dieter Gantenbein, Rainer F. Hauser, Eduard Mumprecht

OSI-Internetworking: Realization and Performance Analysis

In this paper the implementation and performance analysis of an experimental OSI internetwork of interconnected LANs and WANs is described. In the first part, a survey of existing internetworking solutions is given, then the OSI internetwork realized is described. In the second part, methods and tools developed for the performance analysis of heterogeneous networks are introduced. This comprises the simulative modeling system NETSIM and the distributed monitoring system NETMON. These systems are also suitable for the debugging and optimizing implemented code and have been applied to the analysis and optimization of the experimental OSI internetwork.
Martin Zieher, Willi Stoll, Dieter Gantenbein

Protocol Test Tools and Test Methods

This paper addresses the problem of testing OSI protocol implementations. Aspects of the test setup and ways of specifying test cases are discussed. Some experiences made with an experimental test system (KATE) are presented. They show the usefulness of the KATE approach, which was used to test different implementations of the Class 4 OSI Transport Protocol in a heterogeneous environment.
Willi Stoll, Eduard Mumprecht

Database Support for CAM Applications (Project F3)



Object-oriented database systems are one of the more promising concepts for database support for engineering applications. In the literature two levels of object-orientation are distinguished: structural and the behavioral object orientation. A structurally object-oriented database system merely supports the representation of objects by providing expressive data modelling constructs. The behaviorally object-oriented systems go beyond this: Aside from the structural representation they also provide language concepts to model the behavior of objects.
P. Dadam, R. Dillmann, A. Kemper, P. C. Lockemann

Object-Oriented Databases for Robot Programming

R2D2 (Relational Robotics Database System with Extensible Data Types) is a cooperative project that aims at database support for engineering applications in ways that give the engineering user a view of the database which closely models his way of thinking about his world. The approach taken is that of object-orientation. The basic building block of the R2D2 system is a structurally object-oriented database system prototype, the AIM-P system, which offers an extended NF2 relational data model. It is augmented by a behaviorally object-oriented layer of user-definable abstract data types, that adds data selectivity and runtime performance to application programs. The topmost layer contains Pascal programs for engineering applications. In the project the application is off-line simulation of robot motion programs; it is used to validate the general approach. The paper discusses the overall motivation for the project, namely system integration in manufacturing automation, the role of databases in integration, the ensuing challenges to database technology, how these are being met in the project by object-orientation, the proper design of an interface that is easy to use by the application programs, and finally the validation by robot motion simulation.
P. Dadam, R. Dillmann, A. Kemper, P. C. Lockemann

Managing Complex Objects in R2D2

R2D2 — A Relational Robotics Database System with Extensible Data Types — is a joint project of the IBM Heidelberg Scientific Center and the University of Karlsruhe, Computer Science Department. It aims at the design and implementation of a data management system to support engineering (esp. robotics) applications. The management of complex objects in R2D2 is supported by the underlying database system AIM-P. In this paper we describe the AIM-P system in some detail, its data model and database language, and we emphasize the aspects of extensibility by user defined data types and functions.
P. Dadam, K. Küspert, N. Südkamp, R. Erbe, V. Linnemann, P. Pistor, G. Walch

Object Orientation in R2D2

Object-orientation is usually approached from either the programming language or from the database system level. The former emphasizes the application-specific manipulation of technical objects while hiding their structural details whereas the latter concentrates on the structural aspects. The thesis of this paper is that the two viewpoints can be combined to form one homogeneous concept. Thereby we facilitate the manipulation of objects within the application program and the management of persistent objects within the database system. Objects exhibit the same behavior at both system levels, that is within the database kernel as well as within the programming environment. A nested relational database system forms the basis for the object-oriented CAD/CAM programming environment in our system R2D2. Objects defined in the form of abstract data types at the database level can be fetched into an object cache in which they are automatically transformed into a main memory representation. Such a checked-out object is appropriately locked within the database in order to control concuurent transactions that may try to access (and modify) the same objects. The caching technique provides for the required performance of CAD/CAM applications.
Alfons Kemper, Mechtild Wallrath, Martin Dürr

R2D2: An Integration Tool for CIM

Programming of action sequences for multirobot assembly cells involves the representation, the specification and the modelling of the cell activities. For this purpose a chain of activities producing and consuming information has to be performed. Process planning in manufacturing requires that a product’s design must be translated into the best method for the product’s manufacture. The final plan will consist of information about the manufacturing process to be applied, its process parameters, the machines to be used and the time schedule. Synthesizing of the process information and the use of decision logic may be performed by a human expert aided by CAD and CAM software tools or by high sophisticated automatic planner software systems. The bridge between CAD, CAM and Robotics would be the use of a database management system, which supports each activity with appropriate information. A basic representation scheme for the different information and object classes used for planning and programming facilitates the common share of information. In this paper the use of the NF2 data model is discussed to build manufacturing cell models, to construct product assembly plan information and to specify elementary robot actions. The steps for constructing an executable robot assembly program is illustrated. The program is tested with the aid of the graphical Robot Simulation system ROSI.
R. Dillmann, M. Huck

The User Guidance System BFS (Project F4)



The activities of my chair for the Organisation of Data Systems are closely related to the tasks I have as Director of the Computing Centre of the University. The research projects of my chair are always initiated by very practical questions which are highly interesting to the users of our equipment as well as to our students.
A. Schreiner

A System for User Guidance in an Academic Computer Network

In this paper a conception a knowledge based information system, which give users a survey on the activities of the computer centre, equipment, software products, etc., is presented. This system called user guidance system is designed to take up knowledge about different kinds of applications.
The declarativ knowledge is represented in a hierachical class-relationship model describing the structural knowledge concerning the computer centre. Moreover a formal description of the current state of a user request during a consultation is shown using an expansion of this model with restricted and selected relations.
The support of knowledge acquisition for the user guidance system is of particulary importance. Therefore the conception of a knowledge acquisition module is outlined which allows the domain expert (the advisor of the computer centre) to maintain the knowledge base without the help of the system development team.
Finally some aspects of the implementation and the availability of the system are given at the end of this paper.
Gerd Mettendorf, Wilhelm Fries

KESS: A hybrid expert system shell

This paper describes the concept of the hybrid expert system shell KESS (Karlsruher Expert System Shell), the general architecture and the special features in representation, problem solving, explanation and customization.
Johannes Ecke-Schüth
Weitere Informationen