Prashant R. Chandra

Darwin: customizable resource management for value-added network services

The Internet is rapidly changing from a set of wires and switches that carry packets into a sophisticated infrastructure that delivers a set of complex value-added services to end users. Services can range from bit transport all the way up to distributed value-added services like video teleconferencing, virtual private networking, data mining, and distributed interactive simulations. Before such services can be supported in a general and dynamic manner, we have to develop appropriate resource management mechanisms. These resource management mechanisms must make it possible to identify and allocate resources that meet service or application requirements, support both isolation and controlled dynamic sharing of resources across services and applications sharing physical resources, and be customizable so services and applications can tailor resource usage to optimize their performance. The Darwin project has developed a set of customizable resource management mechanisms that support value-added services. We present and motivate these mechanisms, describe their implementation in a prototype system, and describe the results of a series of proof-of-concept experiments.

Network support for application-oriented QoS

Addresses a dilemma raised by recent advances in networking technology, which provide support both for a rich variety of qualities of service (QoSs) and for applications that connect many end-points. Together these features encourage the development of complex multi-party applications that use a diverse set of data types. This raises a two-fold problem: how do application designers choose and specify the many QoS parameters that drive the ultimate performance of their applications; and how does the network efficiently manage its resources to support such a rich application mix? Our approach to this problem is to allow applications to be built around value-added services that encapsulate a variety of simpler resources. This enables both the specification of QoS in terms meaningful to applications, and global optimization of resource allocation across multiple streams and data types. We present a network architecture and a preliminary implementation that explicitly support the notion of application-oriented QoS for complex network services. The key concept is that of service brokers, which applications and service providers use to identify the network resources needed to meet QoS and cost objectives. Service brokers can incorporate a detailed understanding of an application domain, allowing them to make intelligent tradeoffs and to interact with applications and service providers at a high level. They can be hierarchical, in the sense that one broker can invoke the services of another broker. Finally, they provide the ability to deal with heterogeneous networks and hierarchical resource management.

Customizable virtual private network service with QoS

Abstract In this paper, we propose and implement Virtual Network Service (VNS), a value-added network service for deploying virtual private networks (VPNs) in a managed wide-area IP network. The key feature of VNS is its capability of providing a customer with a VPN that is customizable with management capabilities and performance properties comparable to a dedicated physical network. In addition, VNS ensures confidentiality of data and principals through the use of IPSEC. The main technique underlying VNS is the virtualization of routers in both control and data planes. Virtualization of the control plane enables customizable routing and signaling per VPN. On the data plane, packet forwarding and link bandwidth are virtualized. Virtualization of the forwarding mechanism on the data plane enables forwarding of traffic according to each VPNs topology and policies. Virtualization of the link bandwidth enables each VPN to have guaranteed quality of service (QoS) and customized resource management policies. We have developed a VNS prototype for deployment on the CAIRN network. The VNS prototype implements several resource management mechanisms including packet scheduling, signaling and runtime monitoring. A graphical user interface enables service providers to manage, configure and deploy VPNs remotely.

A signaling protocol for structured resource allocation

There is an emerging class of multi-party, multimedia, multi-flow applications that have a high-level structure that imposes dependencies between resource allocations for flows within the application. These applications are also capable of making intelligent decisions on how resource allocation should be controlled within the application. The development of such applications places new requirements on signaling protocols. This paper outlines these new requirements, discusses ways in which they can be supported and presents the design and implementation of an experimental signaling protocol that supports these requirements. This paper makes the case that for these structured applications, there is an advantage to allocating the resources in an integrated fashion, i.e. computation, storage and communication resources for all the flows are allocated at the same time in a coordinated fashion. The concept of a virtual mesh is introduced as a key abstraction that encapsulates the set of resources that are allocated and managed in an integrated fashion to meet the needs of applications. The paper presents two mesh setup algorithms and a performance evaluation comparing them. Temporal resource sharing within the virtual mesh is discussed in detail and signaling support for temporal sharing at setup and runtime is examined. It is important to characterize temporal sharing since it can significantly reduce the resource requirements for applications. We have implemented the Beagle signaling protocol that supports this integrated resource management model. Beagle representations and mechanisms for mesh setup and temporal sharing are described and a prototype implementation is presented.

An active networking approach to service customization

Active networking is a powerful technology to insert new functionality into networking. We look at how active networking technology can be used to customize network services. We observe that users often want slightly different versions of network services such as multicast and network quality of service. We propose to implement these services as a base service that provides the basic service functionality and a customization code module that allows users to customize the service. The customization module uses a service-specific API to modify service behavior. We compare this architecture with the traditional active networking architecture based on execution environments and active applications. We also present several examples of customizable network services.

Implementing Voice over AAL2 on a Network Processor

Publisher Summary The flexibility and programmability of next-generation network-processing units (NPU) make them a key component of next-generation telecommunications equipment. NPUs can support a variety of packet-processing applications, with a variety of different requirements. This chapter discusses the special requirements and challenges presented by the Voice over AAL2 (VoAAL2) application. The architecture and design of a VoAAL2 application developed for the Intel IXP2400 processor is described, and the features of an NPU that are required to support this type of application are discussed. With strict quality of service requirements and asynchronous inputs, VoAAL2 can be performed on an NPU. The application is most naturally implemented using an asynchronous programming model. It is found that the IXP2400 naturally supports such a programming model. Support for asynchronous components that span microengines can be improved by adding support for a distributed lock manager.

Intel Corporation—Intel IXP2400 Network Processor: A Second-Generation Intel NPU

Publisher Summary Next generation access and edge equipment requires flexible programming, high performance, low power consumption, and small real estate. In context to this, Intel has developed a next-generation network processor, the IXP2400, which is optimized to meet these requirements. The Intel IXP2400 network processor delivers a new level of intelligence and performance for access and edge applications, enabling the realization of quality of service (QoS), enforcement of service-level agreements (SLAs), and traffic engineering at OC-48/2 .5 Gbps and 4 Gbps data rates. The flexible media interface allows a variety of media devices, ranging from OC3 to OC48 speeds to be connected without logic to the IXP2400 for easier design and lower system cost. These capabilities essentially allow OEMs and service providers to offer differentiated and tiered services to their customers while efficiently managing their network resources and bandwidth. The performance and flexibility of the IXP2400 makes it desirable for a wide variety of high-performance applications such as multiservice switches, DSLAMs (DSL access multiplexers), CMTS (cable modem termination system) equipment, 2.5G and 3G wireless infrastructure, and layer 4-7 switches.

Extensible signaling for temporal resource sharing

The Internet is rapidly evolving from a network that provides basic best-effort communication service to an infrastructure capable of supporting complex value-added services. These services typically have multiple fluffs with interdependent resource requirements. These dependencies provide opportunities to share the same set of resources among related flows over time leading to significant resource gains. We call this type of sharing temporal resource sharing. Exploiting temporal sharing requires support in the signaling protocol that performs resource allocation for the related flows. We examine the problem of supporting temporal sharing in a signaling protocol. This paper makes the case that temporal sharing support must be designed to be extensible, so that service providers can define and implement new sharing behaviors without having to modify the signaling protocol. We motivate the need for an extensible design by showing that the range of possible temporal sharing behaviors is large and supporting the most general forms of temporal sharing is computationally expensive. We then present a design for extensible signaling support for temporal sharing. We have implemented the temporal sharing design presented in this paper in the Beagle signaling protocol. We present an evaluation of the Beagle design and contrast it with other signaling protocols like RSVP and Tenet-2.

Latency Hiding in Multi-Threading and Multi-Processing of Network Applications

Network processors employ a multithreaded, chip-multiprocessing architecture to effectively hide memory latency and deliver high performance for packet processing applications. In such a parallel paradigm, when multiple threads modify a shared variable in the external memory, the threads should be properly synchronized such that the accesses to the shared variable are protected by critical sections. Therefore, in order to efficiently harness the performance potential of network processors, it is critical to hide the memory latency and synchronization latency in multi-threading and multiprocessing. In this paper, we present a novel program transformation used in the Intelreg Auto-partitioning C Compiler for IXP, which perform optimal placement of memory access instructions and synchronization instructions for effective latency hiding. Experimental results show that the transformation provides impressive speedup (up-to to 8.5x) and scalability (up- to 72 threads) of the performance for the real-world network application (a 10Gbps Ethernet Core/Metro Router).

Latency hiding through multithreading on a network processor

1. IXP Architecture Network processors are specialized processors used to build network devices such as switches, routers, firewalls, etc. because of their flexibility, programmability and ability to deliver scalable packet processing performance from a few 100 Mbps to 10Gbps. However, because of the highly multi-threaded, chipmultiprocessor architectures of network processors, most developers find it difficult to realize their full performance potential for their target applications and often resort to programming in low-level assembly language. The resulting complexity of software development often masks the benefits of employing network processors.