Many real-time application domains can benefit from flexible and open distributed architectures, such as those defined by the CORBA specification. CORBA is an architecture for distributed object computing being standardized by the OMG. Although CORBA is well-suited for conventional request/response applications, CORBA implementations are not yet suited for real-time applications due to the lack of key quality of service (QoS) features and performance optimizations. This paper makes three contributions to the design of realtime CORBA systems. First, the paper describes the design of TAO, which is our high-performance, real-time CORBAcompliant implementation that runs on a range of OS platforms with real-time features including VxWorks, Chorus, Solaris 2.x, and Windows NT. Second, it presents TAO’s realtime scheduling service that can provide QoS guarantees for deterministic real-time CORBA applications. Finally, the paper presents performance measurements that demonstrate the effects of priority inversion and non-determinism in conventional CORBA implementations and how these hazards are avoided in TAO. 1

To be an effective platform for performance-sensitive real-time and embedded applications, off-the-shelf CORBA middleware must preserve the communication-layer quality of service (QoS) properties of applications end-to-end. However, the standard CORBA GIOP/IIOP interoperability protocols are not well suited for applications that cannot tolerate the message footprint size, latency, and jitter associated with general-purpose messaging and transport protocols. It is essential, therefore, to develop standard pluggable protocols frameworks that allow custom messaging and transport protocols to be configured flexibly and used transparently by applications. This paper provides three contributions to research on pluggable protocols frameworks for performance-sensitive distributed object computing (DOC) middleware. First, we outline the key design challenges faced by pluggable protocols developers. Second, we describe how we resolved these challenges by developing a pluggable protocols framewo...

Modern TCP implementations are capable of very high point-to-point bandwidths. Delivered performance on the fastest networks is often limited by the sending and receiving hosts, rather than by the network hardware or the TCP protocol implementation itself. In this case, systems can achieve higher bandwidth by reducing host overheads through a variety of optimizations above and below the TCP protocol stack, given support from the network interface. This paper surveys the most important of these optimizations, and illustrates their effects quantitatively with empirical results from a an experimental network delivering up to two gigabits per second of point-to-point TCP bandwidth. 1 Introduction Good TCP/IP protocol implementations are capable of transferring data at a high percentage of available network link bandwidth, reflecting the success of many years of refinements to TCP/IP protocol handling software and policies. On the fastest networks, applicationto -application throughput is...

To be an effective platform for performance-sensitive real-time systems, distributed object computing middleware must support application quality of service (QoS) requirements end-toend. This article describes how the OMG's Real-time CORBA specification defines standard policies and mechanisms that permit the specification and enforcement of end-to-end QoS.

Distributed object computing forms the basis for nextgeneration application middleware. At the heart of distributed object computing are Object Request Brokers (ORBs), which automate many tedious and error-prone distributed programming tasks. This article presents a case study of key design patterns needed to develop ORBs that can be dynamically configured and evolved for specific application requirements and system characteristics.

Rights to individual papers remain with the author or the author's employer. Permission is granted for noncommercial reproduction of the work for educational or research purposes. This copyright notice must be included in the reproduced paper. USENIX acknowledges all trademarks herein. For more information about the USENIX Association:

Recent advances in I/O bus structures (e.g., PCI), highspeed networks, and fast, cheap disks have significantly expanded the I/O capacity of desktop-class systems. This paper describes a messaging system designed to deliver the potential of these advances for network storage systems including cluster file systems and network memory. We describe gms net, an RPClike kernel-kernel messaging system based on Trapeze, a new firmware program for Myrinet network interfaces. We show how the communication features of Trapeze and gms net are used by the Global Memory Service (GMS), a kernel-based network memory system. The paper focuses on support for zero-copy page migration in GMS/Trapeze using two RPC variants important for peer-peer distributed services: (1) delegated RPC in which a request is delegated to a third party, and (2) nonblocking RPC in which replies are processed from the Trapeze receive interrupt handler. We present measurements of sequential file access from network memory in t...

This paper describes the design and performance of a realtime I/O (RIO) subsystem that supports real-time applications running on off-the-shelf hardware and software. This paper provides two contributions to the study of real-time I/O subsystems. First, it describes how RIO supports end-to-end, prioritized traffic to bound the I/O utilization of each priority class and eliminates the key sources of priority inversion in I/O subsystems. Second, it illustrates how a real-time I/O subsystem can reduce latency bounds on end-to-end communication between high-priority clients without unduly penalizing low-priority and best-effort clients. 1 Introduction This paper focuses on the design and performance of a realtime I/O (RIO) subsystem that enhances the Solaris 2.5.1 kernel to enforce the QoS features of the The ACE ORB (TAO) [1] endsystem. RIO provides QoS guarantees for vertically integrated ORB endsystems in order to (1) increase throughput, (2) decrease latency and (3) improve end-to-end...

Designing ASICs for each new generation of backbone routers is a time intensive and fiscally draining process. In this paper we focus on the design of a programmable architecture for backbone routers, based on the manipulation of wide irregular memory words, that can provide a feasible design alternative to custom ASICs. We propose a pipelined memory design that emphasizes worst-case throughput over latency, and co-explore architectural tradeoffs with the design of several important network algorithms. Through this co-exploration, we show that a programmable architecture can efficiently exploit behavior inherent to most common network algorithms to keep up with next generation network speeds.

A library of layered protocol wrappers has been developed that process Internet packets in reconfigurable hardware. These wrappers can be used with a reprogrammable network platform called the Field Programmable Port Extender (FPX) to rapidly prototype hardware circuits for processing Internet packets. We present a framework to streamline and simplify the development of networking applications that process ATM cells, AAL5 frames, Internet Protocol (IP) packets and UDP datagrams directly in hardware.