As e-business continues to gain momentum, it also develops into more complex and more dynamic forms, both in the B2C (Business to Consumer) sector, and even more so in the B2B (Business to Business) arena.These trends need to be enabled by increased intelligence and capabilities in the network infrastructure and services, which will evolve into an intelligent infrastructure for e-business.Dynamic e-business will enable very rapid establishment of business partnerships and contracts over the global network, without any prior business relation developed in the traditional way.Online business directories, e-markets, and other kinds of e-intermediaries will be deployed to locate suitable partners who can provide desired products, parts, consulting on a specific issue, or financial investment.Trusted e-intermediaries will also offer services to verify the trustworthiness of potential partners and facilitate the negotiation and closure of a contract.All of this will happen at electronic speed, as these processes will be conducted over the network in a highly automated way.Such dynamic processes will be particularly useful in the creation and provisioning of application-level, fully electronic services, or e-services.The providers need to introduce such services to the market quickly and test their market value and acceptance without heavy investments in their own equipment; this need will be addressed by e-utilities, which will become an inseparable part of the intelligent infrastructure.E-utilities will manage thousands of modular servers and storage units that can be allocated on demand to a specific customer, service, or business process, as well as securely released and reallocated as the demand changes.

Network computing is an emerging paradigm that promises to extend computing and information access capabilities transparently beyond the domain of a specific organization and enable unprecedented collaboration and business transactions in a global open network environment. It also challenges the present client-server paradigm, which has proven to be rather expensive, owing to the high cost of maintaining and managing individual workstations and PCs. Network computing is closely coupled with the Internet and World Wide Web explosion. The HTTP-based browser is becoming a standard open universal interface not just for WWW browsing, but for access to corporate databases, for remote system/network management, communication and collaboration, workflow management, and for other services and applications. The Network Computer (NC) is viewed by many as the successor of the familiar PC, that will get most of the needed software and services on demand from the network. To realize the vision of network computing, powerful, robust, and efficient “backend” infrastructure is required to provide such applications and services within an organization, across domains, and on the global scale. What kind of role does the autonomous distributed system (ADS) paradigm play in network computing? Are these two different worlds? Are they complementary? Or, is ADS an inseparable part of or a base for the fledging network computing? The individual panelists address these questions from the point of view of different environments and applications. They discuss technologies that support these paradigms and form a bridge between them.

A review is presented of the different aspects of heterogeneity that need to be resolved to achieve global interoperability, and specifically, to provide global transport service. The authors concentrate on the problems of interconnecting computer networks and systems that use different protocols, either different subsets of OSI (open systems interconnection) standards or OSI and non-OSI protocols. The issues of protocol conversion are discussed, and integrated naming, addressing, and directory mechanisms for interconnecting OSI and non-OSI domains are outlined. More subtle aspects of interoperability are examined for the case of the OSI transport protocol class 4 in mixed LAN/WAN (local-area-network/wide-area network) environments.>

The correct and efficient implementation of complete open systems interconnection systems comprising the specified multiple layers of complex protocols is discussed. Also examined are the specification, design, and implementation issues and approaches at two levels: the realization of the individual protocols; and the construction of complete multilayer communication systems. The design of a multilayer communication subsystem which can run on the same processor as the applications or be confined to a front-end processor is discussed.>

The impact of various implementations of an end-to-end transport service on the performance of high-volume data transfers in a local-area network is investigated. The impact of using a connection-oriented logical-link control (LLC) protocol below a connection-oriented transport protocol is of particular interest. The performance measure is the throughput which can be achieved between two stations either from memory to memory or from disk to disk. The throughput was determined by simulation for various combinations of differing LLC protocols, disk models and processor speeds. The parameters for the simulation model were determined partially by measurements of an implemented file server. The results show that the performance penalty of a connection-oriented LLC protocol can be substantial, but can be significantly lowered when an acknowledgment accumulation strategy is applied.

A control abstraction called atomic action is a powerful general mechanism for ensuring consistent behavior of a system in spite of failures of individual computations running in the system, and in spite of system crashes. However, because of the ``all-or-nothing'' property of atomic actions, an important amount of work might be abandoned needlessly when an internal error is encountered. This paper discusses how implementation of resilient distributed systems can be supported using a combination of nested atomic actions and stable checkpoints. Nested atomic actions form a tree structure. When an internal atomic action terminates, its results are not made permanent until the outermost atomic action commits, but they survive local node failures. Each subtree of atomic actions is recoverable individually. A checkpoint is established in stable storage as part of a remote request so that results of such a request can be reclaimed if the requesting node fails in the meantime, The paper shows how remote procedure call primitives with ``at-most-once'' semantics and recovery blocks can be built with these mechanisms.

This paper describes the design and implementation of filing and printing services in a distributed system based on a token-ring local-area network. The main emphasis is put on the communication aspects of the client/server scenario: roles of a client and a server in a communication protocol, and the integration of communication protocols with applications.

Emulation of systems makes it possible to combine the predictive power of simulation with the advantages of measurement carried under a real system workload. An emulator is a microprogrammed implementation of the basic hardware machine. It can be easily instrumented to collect performance statistics on the instruction set processor (ISP) level and support performance measurement of different configurations and software of the emulated system. This paper describes the monitoring capabilities of the Microprogrammable Multi-Processor (MMP), a powerful emulator system that serves as an experimental tool for evaluating computer systems. The measurement capabilities of the MMP on various system levels are described, as well as existing performance monitoring tools and their applications. Preliminary results contrasting the Gibson mix and measured instruction frequencies on the AN/GYK-12 computer in a TACFIRE system are given.

Two monitors were implemented to collect information about the behavior of the online system developed and run at Stanford. The response of this online system was slow and main memory was a critical resource. The goal was to extract desired information by a method that requires only a negligible amount of monitored system resources. Results presented in this paper indicate that this effort has been successful. A software monitor that requires less than 700 bytes of main memory collects statistics about utilization of special online system resources and about the scheduler mechanism, detects system deadlocks, and measures online executive overhead. This software monitor helped to discover various facts about the system behavior; however, to understand the reasons behind certain situations, it was necessary to learn more about properties of individual terminal tasks. Since a software monitor would cause an intolerable system degradation and hardware monitoring is not directly applicable for such measurements, a hardware/software monitor interface was implemented which enables recording of software events by a hardware monitor. The monitoring artifact is thus kept close to zero. This technique has been applied to measure time a task spends in various states and it has many other uses.