Showing papers in "Ibm Systems Journal in 1995"

The SP2 high-performance switch

Abstract: The heart of an IBM SP2™ system is the HighPerformance Switch, which is a low-latency, highbandwidth switching network that binds together RISC System/6000® processors. The switch incorporates a unique combination of topology and architectural features to scale aggregate bandwidth, enhance reliability, and simplify cabling. It is a bidirectional multistage interconnect subsystem driven by a common oscillator, and delivers both data and service packets over the same links. Switching elements contain a dynamically allocated shared buffer for storing blocked packet flits. The switch is constructed primarily from switching elements (the Vulcan switch chip) and adapters (the SP2 communication adapter). The SP2 communication adapter uses a variety of techniques to improve bandwidth and offload communication tasks from the node processor. This paper examines the switch architecture and presents an overview of its support software.

DB2 parallel edition

Abstract: The rate of increase in database size and response-time requirements has outpaced advancements in processor and mass storage technology. One way to satisfy the increasing demand for processing power and input/output bandwidth in database applications is to have a number of processors, loosely or tightly coupled, serving database requests concurrently. Technologies developed during the last decade have made commercial parallel database systems a reality, and these systems have made an inroad into the stronghold of traditionally mainframe-based large database applications. This paper describes the DB2® Parallel Edition product that evolved from a prototype developed at IBM Research in Hawthorne, New York, and now is being jointly developed with the IBM Toronto laboratory.

Parallel file systems for the IBM SP computers

Abstract: Parallel computer architectures require innovative software solutions to utilize their capabilities. This statement is true for system software no less than for application programs. File system development for the IBM SP product line of computers started with the Vesta research project, which introduced the ideas of parallel access to partitioned files. This technology was then integrated with a conventional Advanced Interactive Executive™ (AIX™) environment to create the IBM AIX Parallel I/O File System product. We describe the design and implementation of Vesta, including user interfaces and enhancements to the control environment needed to run the system. Changes to the basic design that were made as part of the AIX Parallel I/O File System are identified and justified.

The communication software and parallel environment of the IBM SP2

Abstract: This paper describes the software available on the IBM SP2™ for parallel program development and execution. It presents the rationale for the design of the Message-Passing Library used on the SP2, outlines its current implementation, and gives information on performance. In addition, the paper describes the programming environment and the program development tools available for developing and executing parallel codes.

ABC++: concurrency by inheritance in C++

Abstract: Many attempts have been made to add concurrency to C++, often by extensive compiler extensions, but much of the work has not exploited the power of C++. This paper shows how the object-oriented facilities of C++ are powerful enough to encapsulate concurrency creation and control. We have developed a concurrent C++-based prototype system (ABC++) and describe how we can provide, with a standard compiler, almost all of the functionality offered by a new or extended language. Active objects, object distribution, selective method acceptance, and synchronous and asynchronous object interaction are supported. Concurrency control and synchronization are encapsulated at the active object level. The goal of ABC++ is to allow users to write concurrent programs without dealing with explicit synchronization and mutual exclusion constructs, with as few restrictions on the use of C++ as possible. ABC++ can be implemented on either a shared memory multiprocessor or a cluster of homogeneous workstations. It is presently implemented on a network of RISC System/6000® processors and on the IBM Scalable POWERparallel™ System 1 (SP1™).

NBBS traffic management overview

Abstract: In this paper, we describe an integrated set of procedures used for bandwidth management and congestion control in high-speed packet-switched networks such as asynchronous transfer mode (ATM), which are part of IBM's Networking BroadBand Services (NBBS) architecture. These controls are designed to support a wide variety of services with different characteristics in the network and operate at different time scales: connection-level controls such as path selection, admission control and bandwidth allocation, and packet-level controls that discriminate between packets from different connections to support multiple levels of service guarantees. Connection-level controls are applied at connection setup time and are based on the connection characterization and the network state at that time. They perform efficient allocation of resources to ensure performance guarantees for connections while achieving high utilization of network resources. Various packet-level controls developed include access or rate control and intermediate node buffer management and scheduling. For connections that do not require explicit service guarantees, NBBS offers an available bit rate service. This service mostly relies on packet-level control in the form of an end-to-end rate-based flow control algorithm that regulates the flow of traffic into the network. This paper, in addition to providing an overview of the different mechanisms used for traffic management in NBBS, highlights how they interact to ensure efficient network operation.

The CORDS multidatabase project

Abstract: In virtually every organization, data are stored in a variety of ways and managed by different database and file systems. Applications requiring data from multiple sources must recognize and deal with the specifics of each data source and must also perform any necessary data integration. The objective of a multidatabase system is to provide application developers and end users with an integrated view of and a uniform interface to all the required data. The view and the interface should be independent of where the data are stored and how the data are managed. CORDS is a research project focused on distributed applications. As part of this project, we are designing and prototyping a multidatabase system. This paper provides an overview of the system architecture and describes the approaches taken in the following areas: management of catalog information, schema integration, global query optimization, (distributed) transaction management, and interactions with component data sources. The prototype system gives application developers a view of a single relational database system. Currently supported component data sources include several relational database systems, a hierarchical database system, and a network database system.

High-performance parallel implementations of the NAS kernel benchmarks on the IBM SP2

Abstract: Recently, researchers at NASA Ames have defined a set of computational benchmarks designed to measure the performance of parallel supercomputers. In this paper, we describe the parallel implementation of the five kernel benchmarks from this suite on the IBM SP2™, a scalable, distributed memory parallel computer. High-performance implementations of these kernels have been obtained by mapping the computation of these kernels to the underlying architecture of the SP2 machine. Performance results for the SP2 are compared with publicly available results for other high-performance computers.

Local area networks—evolving from shared to switched access

Abstract: Local area networks (LANs) have become pervasive in their application to business, government, and academic data communications. In the United States alone, by 1993 there were over three million LANs installed. This paper describes how advances in structured cabling systems and electronics concentrated in the wiring closet have enabled the evolution of LANs toward a dedicated bandwidth model. Due to the growing number of users, LANs are required to provide increased data capacity and improved management capabilities. These user requirements are being met by increased functions in the electronic wiring closet that include fault isolation, remote network management, increased aggregate bandwidth, and virtual LAN capabilities. This paper focuses on present and future user requirements and the influences of emerging LAN switching technologies on already installed LANs.

NBBS path selection framework

Abstract: This paper describes the path selection function in Networking BroadBand Services (NBBS), which is IBM's architecture for high-speed, multimedia networks. The distinguishing feature of a multimedia network is its ability to integrate different applications with different traffic characteristics and service requirements in the network, such as voice, video, and data. In order to meet their service requirements, it is necessary for the network to provide unique quality-of-service (QOS) guarantees to each application. QOS guarantees, specified as multiple end-to-end performance objectives, translate into path and link constraints in the shortest path routing problem. For a general cost function, shortest path routing subject to path constraints is known to be a nonpolynomial- (NP-) complete problem. The NBBS path selection algorithm, a heuristic solution based on the Bellman-Ford algorithm, has a polynomial order of complexity. The algorithm finds a minimum hop path satisfying an end-to-end delay (or delay variation) constraint, that in most cases also optimizes a load balancing function. To reduce the number of path constraints, other QOS requirements such as packet loss ratio are implemented as a link constraint. The notion of primary and secondary links is used to minimize the long-term overall call blocking probability by dynamically limiting the hop count of a given path. The path selection algorithm developed for point-to-point connections is described first, followed by its extension to the case of point-to-multipoint connections.

Next generation air traffic control automation

Abstract: The automation of air traffic control (ATC) applications has challenged the technologies traditionally used for building and testing large data processing systems. These systems are characterized by complex applications, extensive external and user interfaces, and rapidly changing functional requirements. The Advanced Automation System (AAS) of the Federal Aviation Administration combines high availability, extensibility, and extraordinary functionality into a single distributed system, running at multiple locations. By requirement, the AAS must execute continuously without error, without interruption for upgrades or maintenance, and with the complete trust and confidence of air traffic controllers using it at over twenty centers. Each ATC center uses the same application software coupled with varying quantities of hardware and an extensive base of data to tailor the center to its unique geography and operational procedures. In this paper, we describe the characteristics and architecture of the AAS and focus on key mechanisms of availability and extensibility.

A scalable implementation of the NAS Parallel Benchmark BT on distributed memory systems

Abstract: In this paper, we describe an efficient and scalable implementation of the NAS Parallel Benchmark BT suitable for distributed memory systems such as the IBM Scalable POWERparallel Systems®. After describing the parallelization and data partitioning methods used, we outline some of the optimization steps used to realize good performance on individual processors and to reduce the communication overheads on the IBM SP1™ and SP2™ systems. We present performance results on up to 128 nodes of the SP1, and on the SP2 with wide nodes. We describe the performance on the standard Class A and Class B problem sets. To show the scalability of our parallelization methods, we present the performance of two additional data sets.

Multiprotocol transport networking: eliminating application dependencies on communications protocols

Abstract: The Multiprotocol Transport Networking (MPTN) architecture is a general solution that breaks the binding between distributed applications and communications protocols. The MPTN architecture enables existing applications to run unmodified over any communications protocol. In this paper, we first present the trends in networking that resulted in today's networks supporting multiple communications protocols. Next, we describe the classes of problems this support causes. The MPTN architecture is described and presented as a solution to many of these problems. We also present several alternative solutions to the multiple communications protocol problem and compare them to the MPTN solution. Last, we describe the IBM AnyNet™ family of products that implement the MPTN architecture.

Experiences with object-oriented group support software development

Abstract: This paper describes practical design and implementation experiences gained when creating Group Support Systems (GSS) in a networked personal computer environment. Examples of GSS based on the shared context model and implemented using C, C++, and Actor languages are presented. Graphical user interfaces and multitasking extend traditional methods for supporting group work. An object-oriented communication system is introduced comprised of objects that provide support for all inter- and intraprocessor communications between the GSS applications. Multiple levels of data service are provided to maintain shared data, coordinate user views, and transmit cursor positions in a convenient and efficient manner. The applications presented not only demonstrate the viability of implementing GSS on personal computer-based systems, but also show the ability to develop complex applications in different programming environments that make use of common routines. The unique properties of the object-oriented paradigm greatly facilitate the creation and use of Group Support Systems.

TCP/IP: the next generation

Abstract: After tracing the evolution of Transmission Control Protocol/Internet Protocol (TCP/IP) from its academic and government research heritage to its current commercial use, we survey new directions in TCP/IP. We explore the issues that derive from growth in scale and function, which are prompting the Internet community to assess significant changes in the protocol suite. We relate these issues to the IBM Open Blueprint™. Much as the IBM Personal Computer brought computing to millions of people, the Internet service of the IBM Global Network and IBM's Internet Connection products are bringing networking to millions of people and are changing how customers are transacting business among themselves and with IBM.

Advances in APPN architecture

Abstract: In this paper, we discuss the evolving environment and requirements for the Advanced Peer-to-Peer Networking™ (APPN®) architecture and the accommodation of these changes in basic directory, topology, and configuration services, as well as in application transport capabilities. We use high-performance routing, a recent APPN extension, as an example of the adaptation of the architecture to emerging highspeed communication facilities and the increasing trend to multiprotocol networks. Finally, we discuss some extensions to switched support and network management in APPN and speculate on possible future considerations.

Overview of the NBBS architecture

Abstract: Networking BroadBand Services (NBBS) architecture has been developed by IBM to meet the requirements of high-speed networking. These include addressing problems caused by emerging multimedia applications with extremely high bandwidth requirements and quality-of-service guarantees, as well as those of supporting a variety of existing protocols and interfaces. This paper offers an overview of NBBS and provides insight into the technology and structure of the architecture. Included is a synopsis of the driving forces behind the architecture and a summary of how the challenges of this new environment are addressed.

Multiprotocol networking—a blueprint

Abstract: IBM's Open Blueprint™ is used as a framework for discussing recent developments in communications with particular emphasis on multiprotocol networking. The major parts of the Open Blueprint that pertain to networking are described, indicating advancements in user networks, networking standards, and their underlying technologies. Asynchronous transfer mode is shown to have special significance both for today's multiprotocol networks and tomorrow's multimedia networks.

IBM wireless RF LAN design and architecture

Abstract: This paper describes the medium access control (MAC) protocol and the network integration of the IBM Wireless Radio Frequency (RF) Local Area Network (LAN) product. The MAC protocol is an adaptive, hybrid scheme relying on reservation-based and contention-based methods. It provides sustained high performances, both for light and heavy traffic conditions. Some control services of the wireless channel are also described: security, data compression, and interference management. The network integration is achieved by relying on the industry standards Open Driver Interface (ODI™) and Network Driver Interface Specification (NDIS). Wireless connectivity can either provide an extension of an existing cabled network or result in a stand-alone wireless network.

Intelligent agents extend knowledge-based systems feasibility

Abstract: Episodic classification problems are the transient responsibilities of the knowledge worker—growing and then receding in importance over time Typically, episodic classification problems do not conform to traditional expert system solutions, and they require specialized architectures to offer decision support and increased span of control for those individuals whose task is expediting the problems This paper describes a system for addressing episodic classification problems, giving appropriate functional and technical detail, and continues by illustrating the effectiveness of the system through a case study of use of the system

Efficient transport and distribution of network control information in NBBS

Abstract: With the advances in fiber-based transmission systems that operate at gigabit rates and with the introduction of high-speed networks, the course of communication and computer technologies has changed forever. This change requires that new attention be focused first on the creation, then on the control of networks, which now contain high-speed links and integrate heterogeneous traffic. IBM's Networking BroadBand Services (NBBS) architecture has been designed to enable this networking revolution and, in particular, is designed for the high-speed, multimedia networks needed by emerging applications. In this paper, we present the Rapid Transport Protocol (RTP). Its simple and efficient mechanisms enable NBBS control information to be transported and distributed, taking advantage of high-speed links by eliminating as much nodal processing as possible. RTP provides point-to-point and point-to-multipoint transport services with a reliable delivery option. In addition, we present a simple and efficient mechanism for fast dissemination of time-critical network configuration and path update messages to every node in an NBBS network, making use of RTP.

NBBS network management

Abstract: Network management for Networking BroadBand Services (NBBS), a general-purpose, multiprotocol, high-speed transport backbone is discussed. The paper begins with the objectives and the key choices made in the development of the network management solution. The managed environment for high-speed transport networks and some features of the NBBS architecture that have influenced the network management solution are described, along with the relationship between the network manager and the managed target system agent, the actual communication environment for the Nways™ Switch Manager. The relationship between the Network Management Workstation and an NBBS node acting as a network management agent is presented. The use of managed object classes to provide a Management Information Base (MIB) that may be accessed by a manager using the Common Management Information Protocol (CMIP) to manage NBBS networks is then discussed. Management application models for accounting, configuration, topology, performance, and fa ult management for NBBS are illustrated, showing how they relate to the NBBS managed objects. The paper concludes with a look at future directions for the NBBS management solution.

Collecting product-based usability requirements

Abstract: Customer requirements play an essential role in product development. Accurate definition and assessment of customer needs and wants form the basis for product offerings. Functional requirements have always been central to product definition, and several methods are available to determine required function. Emphasis on product usability and the end-user interface is continuing to increase as computer products are more widely used. This paper describes a requirements-gathering methodology that focuses on usability and user requirements. The Customer Requirements and Task Specification method was developed over the last six years to meet the growing need for more precise user requirements. Each of the major steps is discussed, as is the role of computer-supported cooperative work techniques. Summary findings from a broad cross section of customers are presented.

Access services for the Networking Broadband Services architecture

Abstract: This paper describes how Networking BroadBand Services (NBBS) provides multiple protocol support over emerging high-bandwidth networks. It describes the problems associated with handling these multiple protocols that are seen now more than ever in the emerging asynchronous transfer mode (ATM) infrastructure. The paper addresses the access services of NBBS and how the NBBS transport functions are used to provide these access services. A detailed example of one access service component is given to provide the reader with a feeling for typical access services issues. Brief examples of access services are given to show how the concepts described are being used.

A real-time systems context for the framework for information systems architecture

Abstract: In this paper we review the framework for information systems architecture first introduced by Zachman 1 and show how it can be applied in the context of real-time systems. Discussions are included throughout the paper to convey some of the characteristics unique to real-time systems and to point out areas of special architectural concern.

ATM: paving the information superhighway

Abstract: A new generation of networking requirements is fueling the growth of a cell-based communications technology known as the asynchronous transfer mode (ATM). ATM technology allows the integration of voice, video, traditional data, and other traffic types on a single network. ATM offers a unique opportunity to deploy the same standardized networking technology in both the wide-area and local-area environments. IBM has introduced a family of products that provides a complete ATM solution for customers. The products support ATM standards, allowing the products to interwork with devices from other vendors. This paper provides a tutorial on ATM technology and an overview of the IBM ATM product family. The IBM 8260 Intelligent Switching Hub is described as a representative ATM product.

An architecture for multimedia communication and real-time collaboration

Abstract: The paper presents the requirements for real-time multimedia communication in a collaborative environment and describes how the requirements can be met through the IBM Lakes architecture. The initiatives of the ITU-T through the T.120 series of recommendations are described, and the interoperability of Lakes with these recommendations is discussed.

Multicast network connection architecture

Abstract: IBM Networking BroadBand Services (NBBS) is a comprehensive networking architecture designed for high-speed networks. To meet the demands of multicast communication applications, e.g., video on demand, software distribution, and video conferencing, NBBS multicast services provide support for efficient best-effort and guaranteed quality-of-service (QOS) multipoint network connections. The NBBS multicast services consist of set management services and multicast network connection services. Set management services provide the ability to create and manage groups of users who are interested in participating in multicast activities. Multicast network connection services use the group membership information provided by set management services to support multiple network connections per group and to provide supporting services such as nondisruptive path switching, path preemption, and bandwidth management. The multicast network connection services support the multicast signaling requirements defined by the asynchronous transfer mode user-to-network interface (ATM UNI) specifications. This paper gives an overview of the NBBS multicast connection services. Set concepts are introduced, an overview of set management protocols is given, and the value of set management services in establishing multicast network connections is explained. Finally, multicast network connection protocols that establish, maintain, and terminate multicast network connections are described.

Experiences with the IBM SP1

Abstract: One of the first IBM parallel processing computers—the SP1™—and the largest, with 128 nodes, was installed in 1993 at Argonne National Laboratory. It took only days, not months, to prepare for and migrate applications to this parallel supercomputer, demonstrating that high performance, parallelism, and portability can coexist. This paper describes the early experiences with the SP1 at Argonne, which provide lessons for supercomputer system designers and users alike. We explore what features of software technology and system architecture enabled immediate and successful use of the SP1. The paper concludes with a brief indication of why the move to the SP2™ software environment using the SP2 communication adapters, the use of the emerging MessagePassing Interface standard, and the continued use of the SP1 processors have been successful.

CPI-C: an API for distributed applications

Abstract: Distributed applications have fostered the standardization of application programming interfaces for the underlying communication services. Three popular communication models—remote procedure calls, messaging and queuing, and conversations—support distributed applications across different networking protocols and physical media. Access to the conversational services of Advanced Program-to-Program Communication and Open Systems Interconnection-Distributed Transaction Processing is provided by the Common Programming Interface for Communications (CPI-C), a standard, easy-to-use interface for communication programming. This paper introduces the basic concepts of CPI-C, describes the conversation services available to programs, and presents examples of CPI-C programming.