Today's computing environment is changing, with growing emphasis on reducing both operational energy costs and the high capital costs to reliably deliver power and cooling to systems in factory-sized data centers. The greening of data centers is challenging microprocessor and system designs, including server, storage, and network designs, to deliver ever higher performance within strict power and cooling constraints. This paper outlines some of the challenges, research approaches and holistic solutions to more energy efficient data centers, starting with microprocessor and system design and extending to the broader environment.

In order to stem the increasing packet loss rates caused by an exponential increase in network traffic, the IETF has been considering the deployment of active queue management techniques such as RED (random early detection) (see Floyd, S. and Jacobson, V., IEEE/ACM Trans. Networking, vol.1, p.397-413, 1993). While active queue management can potentially reduce packet loss rates in the Internet, we show that current techniques are ineffective in preventing high loss rates. The inherent problem with these algorithms is that they use queue lengths as the indicator of the severity of congestion. In light of this observation, a fundamentally different active queue management algorithm, called BLUE, is proposed, implemented and evaluated. BLUE uses packet loss and link idle events to manage congestion. Using both simulation and controlled experiments, BLUE is shown to perform significantly better than RED, both in terms of packet loss rates and buffer size requirements in the network. As an extension to BLUE, a novel technique based on Bloom filters (see Bloom, B., Commun. ACM, vol.13, no.7, p.422-6, 1970) is described for enforcing fairness among a large number of flows. In particular, we propose and evaluate stochastic fair BLUE (SFB), a queue management algorithm which can identify and rate-limit nonresponsive flows using a very small amount of state information.

This paper evaluates techniques for improving operating system and network protocol software support for high-performance World Wide Web servers. We study approaches in three categories: i.e., new socket functions, per-byte optimizations, and per-connection optimizations. We examine two proposed socket functions, i.e., acceptex( ) and send-file( ), comparing send-file( )'s effectiveness with a combination of mmap( ) and writev( ). We show how send-file( ) provides the necessary semantic support to eliminate copies and checksums in the kernel, and quantify the benefit of the function's header and close options. We also present mechanisms to reduce the number of packets exchanged in an HTTP transaction, both increasing server performance and reducing network utilization, without compromising interoperability. Results using WebStone show that our combination of mechanisms can improve server throughput by up to 64%, and can eliminate up to 33% of the packets in an HTTP exchange. Results with SURGE show an aggregate increase in server throughput of 25%.

We propose a flow-control scheme for multicast ABR services in ATM networks. At the heart of the proposed scheme is an optimal second-order rate control algorithm, called the /spl alpha/-control, designed to deal with the variation in RM-cell round-trip time (RTT) resulting from dynamic drift of the bottleneck in a multicast tree. Applying two-dimensional rate control, the proposed scheme makes the rate process converge to the available bandwidth of the connection's most congested link sensed by the traffic source. It also confines the buffer occupancy to a target regime bounded by a finite buffer capacity as the system enters the equilibrium state. It works well irrespective of the topology of the multicast tree. Using the fluid analysis, we model the proposed scheme and analyze the system dynamics for multicast ABR traffic. We study the convergence properties and derive the optimal-control conditions for the /spl alpha/-control. The analytical results show that the scheme is stable and efficient in the sense that both the source rate and bottleneck queue length rapidly converge to a small neighborhood of the designated operating point. We present simulation results which verify the analytical observations. The simulation experiments also demonstrate the superiority of the proposed scheme to the other schemes in dealing with RM-cell RTT and link-bandwidth variations, achieving fairness in both buffer and bandwidth occupancies, and enhancing average throughput.

This paper describes and evaluates stochastic fair blue (SFB), a novel technique for enforcing fairness among a large number of rows. SFB scalably detects and rate-limits non-responsive flows through the use of a marking probability derived from the blue queue management algorithm and a Bloom (1970) filter. Using analysis and simulation, SFB is shown to effectively handle non-responsive flows using an extremely small amount of state information.

This paper exposes an inherent weakness in current active queue manage ment techniques such asRedin that they rely on queue lengths to indicate the severity of congestion. In light of this observation, a fundamentally different active queue management algorithm calledBlueis proposed.Blueuses packet loss and link utilization to manage congestion. Using simulation and controlled experiments,Blueis shown to significantly outperformRedin providing lower packet loss rates and smaller queuing delays to networked applications such as interactive audio and video.

No abstract available.

The Internet provides no support for privacy or authentication of multicast packets. However, an increasing number of applications require secure multicast services in order to restrict group membership and enforce accountability of group members. A major problem associated with the deployment of secure multicast delivery services is the scalability of the key distribution protocol. This is particularly true with regard to the handling of group membership changes, such as member departures and/or expulsions, which necessitate the distribution of a new session key to all the remaining group members. As the frequency of group membership changes increases, it becomes necessary to reduce the cost of key distribution operations. This paper explores the use of batching of group membership changes to reduce the frequency, and hence the cost, of key re-distribution operations. It focuses explicitly on the problem of cumulative member removal and presents an algorithm that minimizes the number of messages required to distribute new keys to the remaining group members. The algorithm is used in conjunction with a new multicast key management scheme which uses a set of auxiliary keys in order to improve scalability. In contrast to previous schemes which generate a fixed hierarchy of keys, the proposed scheme dynamically generates the most suitable key hierarchy by composing different keys. Our cumulative member removal algorithm uses Boolean function minimization techniques, and outperforms all other schemes known to us in terms of message complexity.

The congestion control mechanisms used in TCP have been the focus of numerous studies and have undergone a number of enhancements. However, even with these enhancements, TCP connections still experience alarmingly high loss rates, especially during times of congestion. To alleviate this problem, the IETF is considering active queue management mechanisms, such as random early detection (RED), for deployment in the network. In this paper, we first show that the effectiveness of RED depends, to a large extent, on the appropriate parameterization of the RED queue. We then show that there is no single set of RED parameters that work well under different congestion scenarios. In light of this observation, we propose and experiment with more adaptive RED gateways which self-parameterize themselves based on the traffic mix. The results show that traffic cognizant parameterization of RED gateways can effectively reduce packet loss, while maintaining high link utilizations under a range of network loads.

We propose a flow control scheme for multicast ABR services in ATM networks. At the heart of the proposed scheme is an optimal second-order rate control algorithm, called the /spl alpha/-control, designed to deal with the variation in RM-cell round-trip time (RTT) resulting from dynamic "drift" of the bottleneck in a multicast tree. Applying two-dimensional rate control, the proposed scheme makes the rate process converge to the available bandwidth of the connection's most congested link. It also confines the buffer occupancy to a target regime bounded by a finite buffer capacity. It works well irrespective of the topology of the multicast tree. Using the fluid approximation, we model the proposed scheme and analyze the system dynamics for multicast ABR traffic. We study the convergence properties and derive the optimal control conditions for the /spl alpha/-control. The analytical results show that the scheme is stable and efficient in the sense that both the source rate and bottleneck queue length rapidly converge to a small neighborhood of the designated operating point. We present simulation results which verify the analytical observations. The simulation results also demonstrate the effectiveness of the proposed scheme in dealing with RM-cell RTT and link-bandwidth variations, and in achieving fairness in both buffer and bandwidth occupancies.

No abstract available.

This paper examines the use of adaptive priority marking for providing soft bandwidth guarantees in a differentiated-services Internet. In contrast to other proposals for achieving the same objective, the proposed scheme does not require resource reservation for individual connections and can be supported with minimal changes to the network infrastructure. It uses modest support from the network in the form of priority handling for appropriately marked packets, and relies on intelligent transmission control mechanisms at the edges of the network to achieve the desired throughput levels. This paper describes the control mechanisms and evaluates their behavior in various network environments. These mechanisms are show in to have several salient features which make them suitable for deployment in an evolving Internet.

A large number of Internet applications are sensitive to overload conditions in the network. While these applications have been designed to adapt somewhat to the varying conditions in the Internet, they can benefit greatly from an increased level of predictability in network services. We propose minor extensions to the packet queueing and discard mechanisms used in routers, coupled with simple control mechanisms at the source that enable the network to guarantee minimal levels of throughput to different sessions while sharing the residual network capacity in a cooperative manner. The service realized by the proposed mechanisms is an interpretation of the controlled-load service being standardized by the Internet Engineering Task Force. Although controlled-load service can be used in conjunction with any transport protocol, our focus in this paper is on understanding its interaction with Transmission Control Protocol (TCP). Specifically, we study the dynamics of TCP traffic in an integrated services network that simultaneously supports both best-effort and controlled-load sessions. In light of this study, we propose and experiment with modifications to TCP's congestion control mechanisms in order to improve its performance in networks where a minimum transmission rate is guaranteed. We then investigate the effect of network transients, such as changes in traffic load and in service levels, on the performance of controlled-load as well as best-effort connections. To capture the evolution of integrated services in the Internet, we also consider situations where only a selective set of routers are capable of providing service differentiation between best-effort and controlled-load traffic. Finally, we show how the service mechanisms proposed here can be embedded within other packet and link scheduling frameworks in a fully evolved integrated services Internet.

This paper examines the use of adaptable priority marking for providing soft bandwidth guarantees to individual connections or connection groups over the Internet. In contrast to other proposals for service differentiation which focus on providing firm performance guarantees, the proposed scheme does not require resource reservation for individual connections and can be supported with minimal changes to the network infrastructure. It uses modest support from the network in the form of priority handling for appropriately marked packets and relies on intelligent transmission control mechanisms at the edges of the network to achieve the desired throughput levels. The paper describes the control mechanisms and evaluates their behaviour in various network environments. These mechanisms are shown to have several salient features which make them suitable for deployment in an evolving Internet.

This paper explores the performance impact of supporting QoS guarantees on communication in TCP/IP protocol stacks at Unix-like end hosts. We first demonstrate the efficacy of our RSVP-based QoS architecture in providing the desired QoS to individual connections via application-level experiments using UDP sessions and TCP connections on an ATM network. We then identify and measure, via detailed profiling, the overheads imposed by the individual components of the QoS architecture, such as traffic policing, traffic shaping, and buffer management. Our measurements reveal that traffic policing overheads are largely offset by savings due to per-session buffer pre-allocation, and, for ATM networks, a faster path through the network interface layer. In the latter case the data path latency for compliant packets can even be slightly smaller than the default best-effort data path latency. Traffic shaping presents more challenges, primarily because of interactions with the operating system CPU scheduler. We discuss the performance implications of traffic shaping and suggest techniques to reduce or mask some of the associated overheads.

The Internet Engineering Task Force (IETF) is currently in the process of overhauling the architecture of the Internet to meet new challenges and support new applications. One of the most important components of that venture is the enhancement of the Internet service model from a classless best effort service architecture to an integrated services architecture supporting a multitude of classes and types of services. This paper presents the design, implementation, and experiences with a protocol architecture for the integrated services Internet. It is based on the emerging standards for resource reservation in the Internet, namely, the RSVP protocol and the associated service specifications defined by the IETF. Our architecture represents a major functional enhancement to the traditional TCP/IP protocol stack. It is scalable in terms of performance and number of network sessions, and supports a wide variety of network interfaces ranging from legacy LAN interfaces, such as Token Ring and Ethernet, to high-speed ATM interfaces. The paper also describes the implementation of this architecture on the IBM AIX platform and our experiences with the system. We then present a performance analysis of the system which quantifies the overheads imposed by all components of the QoS support, such as traffic policing, traffic shaping, and buffer management.

An important challenge for the future growth of the Internet is to design routers that can forward the exponentially increasing volume of traffic, and at the same time provide the service differentiation needed by new applications. In this paper, we describe the architecture, implementation, and initial experiences with a system designed to meet this challenge. This system, which we call a QoS capable switch-router (QSR), combines the salient features of switching and routing technologies to provide high throughput and support the different classes of service being defined by the IETF. It consists of a core (ATM) switch fabric connecting intelligent adapters, each capable of both routing and switching pockets. A control engine is responsible for routing, RSVP signalling, and resource management. We have built a prototype network of 3 systems connected to several UNIX hosts, and have conducted preliminary performance measurements on this network.

The paper presents the design and implementation of a quality of service architecture for the Internet. The architecture is based on the emerging standards for resource reservation in the Internet, namely the RSVP protocol and the associated service specifications defined by the Internet Engineering Task Force. Our architecture represents a major functional enhancement to the traditional sockets based communication subsystem, while preserving application programming interface and binary compatibility with existing applications. It is scalable and supports a variety of network interfaces ranging from legacy LAN interfaces, such as token ring and Ethernet, to high speed ATM interfaces. We also describe our initial experiences with the implementation of this architecture on the IBM AIX platform.

At the data-link layer, ATM offers a number of features, such as high-bandwidth and per-session quality-of-service (QoS) guarantees, making it particularly attractive to multimedia applications. Unfortunately, many of these features are not visible to applications because of the inadequacies of existing higher-level protocol architectures. Although there is considerable effort underway to tune these protocols for ATM networks, we believe that a new ATM-specific protocol stack is essential to effectively exploit all the benefits of ATM. We describe the semantics of such a protocol stack, and discuss its advantages over traditional protocol architectures from the perspective of multimedia applications. The performance impact of the new protocol architecture is experimentally demonstrated on a video conferencing testbed built around IBM RS/6000's equipped with prototype hardware for video/audio processing, and connected via ATM links.

We examine two key aspects of the ATM deployment in the TCP/IP environment-logical IP sub-networks (LISs) and address mapping servers, with a specific emphasis on how different alternatives for LISs and address mapping servers enable efficient exploration of all the capabilities that ATM promises to deliver. We propose an approach for ATM deployment in the TCP/IP environment that leverages the existing technologies (e.g. DNS, routing protocols), minimizes new development, imposes no new requirements on ATM, allows for small, incremental changes with immediate payoffs, while at the same time retains backward compatibility with the existing schemes, such as LAN Emulation and "classical" IP over ATM. The proposed approach combines a switch-based infrastructure with router-based overlay and uses each for that which it is best suited: switch-based infrastructure for applications that can justify an SVC establishment; router-based overlay for all other applications.

This paper examines the issues involved in the design of conference servers that support multiparty, multimedia conferences. These servers, called multipoint control units (MCUs) in the telephony world, coordinate the distribution of audio, video, and data streams amongst the multiple participants in a videoconference. The MCU is responsible for the processing of video and audio so that a conference participant can hear and see one or more of the other participants in the conference. It is also responsible for handling and forwarding the data streams from the participants. This paper presents different approaches to the design of an MCU to implement these functions. It also describes the design of a related device-a transcoding gateway that enables conferencing between participants using different video/audio equipment.>

In a video--on--demand (VOD) system, it is desirable to provide the user with interactive browsing functions such as "fast forward" and "fast backward." However, these functions usually require a significant amount of additional resources from the VOD system in terms of storage space, retrieval throughput, network bandwidth, etc. Moreover, prevalent video compression techniques such as MPEG impose additional constraints on the process since they introduce inter-frame dependencies. In this paper, we devise methods to support variable rate browsing for MPEG-like video streams and minimize the additional resources required. Specifically, we consider retrieval for a disk-array-based video server and address the problem of distributing the retrieval requests across the disks. Our overall approach for interactive browsing comprises (1) a storage method, (2) placement and sampling methods, and (3) a playout method, where the placement and sampling methods are two alternatives for video segmen...

Communication in real-time systems has to be predictable, because unpredictable delays in the delivery of messages can adversely affect the execution of tasks dependent on these messages. We develop a scheme for providing predictable interprocess communication in real-time systems with (partially connected) point-to-point interconnection networks, which provide guarantees on the maximum delivery time for messages. This scheme is based on the concept of a real-time channel, a unidirectional connection between source and destination. A real-time channel has parameters that describe the performance requirements of the source-destination communication, e.g., from a sensor station to a control site. Once such a channel is established, the communications subsystem guarantees that these performance requirements will be met. We concentrate on methods to compute guarantees for the delivery time of messages belonging to real-time channels. We also address problems associated with allocating buffers for these messages and develop a scheme that preserves delivery time guarantees.>

A buffer insertion ring with network access fairness control for gigabit LANs is proposed. The basic buffer insertion ring has the maximum spatial reuse on the multiple ring segments by allowing concurrent traffic, but it suffers from the possibility of station starvation. To solve this problem, a notion of adjustable fairness based on mutual interference between pairs of stations in bandwidth usage is defined. The adjustable fairness definition serves as a basis for conditions to trigger backpressure signals. A cooperative backpressure mechanism is invented for flow regulation. The effectiveness of the proposed network access mechanism is demonstrated by comparisons with other buffer insertion rings based on simulations.>

Grouped Sweeping Scheduling was proposed in [1] as a general formulation of a class of disk arm scheduling schemes.This class includes, for example, the fixed-order and SCAN scheduling schemes.An optimum design was presented for the homogeneous case, i.e., when all multimedia streams have the same characteristics.In this paper we examine the more general situation in which a mixture of different format multimedia streams, with different characteristics, coexist on the disk.The emphasis is to present and prove a simple procedure for optimizing GSS in this heterogeneous situation.Finally, we discuss how GSS can be used in dynamic settings to support heterogeneous request streams.

The problem of selecting routes for interprocess communication in a network with virtual cut-through capability while balancing the network load and minimizing the number of times that a message gets buffered is addressed. The approach taken is to formulate the route selection problem as a minimization problem, with a link cost function that depends upon the traffic through the link. The form of this cost function is derived based on the probability of establishing a virtual cut-through route. It is shown that this route selection problem is NP-hard, and an approximate algorithm is developed which tries to incrementally reduce the cost by rerouting traffic. The performance of this algorithm is evaluated for the hypercube and the C-wrapped hexagonal mesh, example networks for which virtual cut-through switching support has been developed.>

A scheme is developed for providing predictable interprocess communication in real-time systems with (partially connected) point-to-point interconnection networks, which provides guarantees on the maximum delivery time for messages. This scheme is based on the concept of a real-time channel, a unidirectional connection between source and destination. A real-time channel has parameters which describe the performance requirements of the source-destination communication, e.g., from a sensor station to a control site. Methods to compute guarantees for the delivery time of messages belonging to real-time channels are examined. Problems associated with allocating buffers for these messages are addressed, and a scheme which preserves delivery time guarantees is developed.>

article Free Access Share on Reliable broadcast algorithms for HARTS Authors: Dilip D. Kandlur Univ. of Michigan, Ann Arbor Univ. of Michigan, Ann ArborView Profile , Kang G. Shin Univ. of Michigan, Ann Arbor Univ. of Michigan, Ann ArborView Profile Authors Info & Claims ACM Transactions on Computer SystemsVolume 9Issue 4pp 374–398https://doi.org/10.1145/118544.119233Published:01 November 1991Publication History 21citation378DownloadsMetricsTotal Citations21Total Downloads378Last 12 Months15Last 6 weeks6 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

The problem of selecting routes for interprocess communication in a network with virtual cut-through capability, while balancing the network load and minimizing the number of times that a message gets buffered, is addressed. The approach taken is to formulate the route selection problem as a minimization problem, with a link cost function that depends on the traffic through the link. The form of this cost function is derived on the basis of the probability of establishing a virtual cut-through route. It is shown that this route selection problem is NP-hard, and so an approximate algorithm that tries to reduce the cost incrementally by rerouting traffic is developed. The performance of this algorithm is evaluated for two popular network topologies: the hypercube and the C-wrapped hexagonal mesh.>

A family of six-regular graphs, called hexagonal meshes or H-meshes, is considered as a multiprocessor interconnection network. Processing nodes on the periphery of an H-mesh are first wrapped around to achieve regularity and homogeneity. The diameter of a wrapped H-mesh is shown to be of O(p/sup 1/2/), where p is the number of nodes in the H-mesh. An elegant, distributed routing scheme is developed for wrapped H-meshes so that each node in an H-mesh can compute shortest paths from itself to any other node with a straightforward algorithm of O(1) using the addresses of the source-destination pair only, i.e. independent of the network's size. This is in sharp contrast with those previously known algorithms that rely on using routing tables. Furthermore, the authors also develop an efficient point-to-point broadcasting algorithm for the H-meshes which is proved to be optimal in the number of required communication steps. The wrapped H-meshes are compared against some other existing multiprocessor interconnection networks, such as hypercubes, trees, and square meshes. The comparison reinforces the attractiveness of the H-mesh architecture.>

A clock synchronization scheme that strikes a balance between hardware and software solutions is proposed. The proposed scheme is a software algorithm that uses minimal additional hardware to achieve reasonably tight synchronization. Unlike other software solutions, the guaranteed worst-cast skews can be made insensitive to the maximum variation of message transit delay in the system. The scheme is particularly suitable for large partially connected distributed systems with topologies that support simple point-to-point broadcast algorithms. Examples of such topologies include the hypercube and the mesh interconnection structures.>