Ashish Mehra

Design and implementation of an RSVP-based quality of service architecture for integrated services Internet

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.

ARMADA Middleware and Communication Services

Real-time embedded systems have evolved during the past several decades from small custom-designed digital hardware to large distributed processing systems. As these systems become more complex, their interoperability, evolvability and cost-effectiveness requirements motivate the use of commercial-off-the-shelf components. This raises the challenge of constructing dependable and predictable real-time services for application developers on top of the inexpensive hardware and software components which has minimal support for timeliness and dependability guarantees. We are addressing this challenge in the ARMADA project.ARMADA is set of communication and middleware services that provide support for fault-tolerance and end-to-end guarantees for embedded real-time distributed applications. Since real-time performance of such applications depends heavily on the communication subsystem, the first thrust of the project is to develop a predictable communication service and architecture to ensure QoS-sensitive message delivery. Fault-tolerance is of paramount importance to embedded safety-critical systems. In its second thrust, ARMADA aims to offload the complexity of developing fault-tolerant applications from the application programmer by focusing on a collection of modular, composable middleware for fault-tolerant group communication and replication under timing constraints. Finally, we develop tools for testing and validating the behavior of our services. We give an overview of the ARMADA project, describing the architecture and presenting its implementation status.

Resource management for real-time communication: making theory meet practice

This paper focuses on bridging the gap between theory and practice in the management of host CPU and link resources for real-time communication. Using our implementation of real-time channels, a paradigm for real-time communication in packet-switched networks, we illustrate the tradeoff between resource capacity and channel admissibility, which determines the number and type of real-time channels that can be accepted for service and the performance delivered to best-effort traffic. We demonstrate that this tradeoff is affected significantly by the choice of implementation paradigms and the grain at which CPU and link resources are multiplexed amongst active channels. To account for this effect we extend the admission control procedure for real-time channels originally proposed using idealized resource models. Our results show that practical considerations significantly reduce channel admissibility compared to idealized resource models. Further, the optimum choice of multiplexing grain depends on several factors such as resource preemption overheads, the relationship between CPU and link bandwidth, and the interaction between link bandwidth allocation and CPU bandwidth allocation.

Design and implementation of an RSVP-based quality of service architecture for an integrated services Internet

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.

The END: a network adapter design tool

We present the design and implementation of the Emulated Network Device (END), a network adapter design tool that facilitates accurate evaluation of alternative adapter designs. Using device emulation, the END permits designers to couple a representative model of an adapter with a real host and its communication software. Different adapter designs can be evaluated and compared accurately in a realistic setting, i.e., while capturing host-adapter concurrency and interaction overheads, before building a prototype. We present the architectural framework adopted by the END and demonstrate its feasibility via a case study of a commercial network adapter. Several design improvements to alleviate performance bottlenecks are realized and evaluated using the END, highlighting its utility as a network adapter design tool.

SPIDER: flexible and efficient communication support for point-to-point distributed systems

SPIDER is a network adapter that provides scalable communication support for point-to-point distributed systems. The device exports an efficient interface to the host processor, provides transparent support for dependable, time-constrained communication, and handles packet routing and switching. The communication support provided by SPIDER exploits concurrency between the independent data channels feeding the point-to-point network, and offers flexible and transparent hardware mechanisms. SPIDER allows the host to exercise fine-grain control over its operation, enabling the latter to monitor and influence data transmission and reception efficiently. In the current implementation, SPIDER interfaces to the Ironics IV-3207, a VMEbus-based 68040 card and will be controlled by x-kernel, a communication executive allowing the flexible composition of communication protocols.<<ETX>>

Exploring the performance impact of QoS support in TCP/IP protocol stacks

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.

Window-consistent replication for real-time applications

Two widely-studied approaches for structuring fault-tolerant services are the state-machine and the primary-backup replication schemes. For a large class of soft and hard real-time applications, the degree of consistency among servers can be exploited to design replication protocols with predictable timing behavior. This is particularly useful in applications such as automated process control, in which one can trade off the quality or precision for timely availability of data. This paper presents the architecture and prototype implementation of a primary-backup replication service that employs window consistency semantics between the primary data repository and the backups. A client registers a data object with the service by declaring the consistency requirements for the data, in terms of a time window. The primary ensures that each backup site maintains a version of the object that was valid on the primary within the preceding time window by scheduling update messages to the backups. Decoupling the transmission of updates to the backups from the processing of client requests permits the primary to handle a higher rate of operations and provide more timely service to clients. The non-blocking semantics free the client from waiting for updates to the backups to complete. Furthermore, real-time scheduling of update messages can guarantee controlled inconsistency between the primary and backup repositories.<<ETX>>

Design and evaluation of a window-consistent replication service

Real-time applications typically operate under strict timing and dependability constraints. Although traditional data replication protocols provide fault tolerance, real-time guarantees require bounded overhead for managing this redundancy. This paper presents the design and evaluation of a window-consistent primary-backup replication service that provides timely availability of the repository by relaxing the consistency of the replicated data. The service guarantees controlled inconsistency by scheduling update transmissions from the primary to the backup(s); this ensures that client applications interact with a window-consistent repository when a backup must supplant a failed primary. Experiments on our prototype implementation, on a network of Intel-based PCs running RT-Mach, show that the service handles a range of client loads while maintaining bounds on temporal inconsistency.

Policy-based Diffserv on Internet servers: the AIX approach

The Internet infrastructure must evolve from best-effort service to meet the needs of different customers and applications. With Internet traffic differentiation, service providers can support a range of offerings, such as loss or delay bounds and network bandwidth allocation, to meet different performance requirements. The differentiated services (Diffserv) architecture provides a scalable approach, in which network access (or edge) devices aggregate traffic flows onto provisioned pipes that traverse a streamlined network core. We have identified the key requirements for provisioning Diffserv functions on Internet servers. Based on these requirements, we have implemented, and deployed, a policy-based architecture on IBMs AIX operating system that provides Diffserv services to both QoS-aware and -unaware applications.