Learn More
In this paper, we describe the design and implementation of an integrated architecture for cache systems that scale to hundreds or thousands of caches with thousands to millions of users. Rather than simply try to maximize hit rates, we take an end-to-end approach to improving response time by also considering hit times and miss times. We begin by studying(More)
In this paper, we examine several distributed caching strategies to improve the response time for accessing data over the Internet. By studying several Internet caches and workloads, we derive four basic design principles for large scale distributed caches: (1) minimize the number of hops to locate and access data, (2) do not slow down misses, (3) share(More)
817 Abstrue&-Thh paper presents detailed measurements of processing overheads for the Ultrix 4.2a implementation of TCP/fP network software running on a DECstat30n 5000/200. The performance results were used to uncover throughput and latency bottlenecks. We present a scheme for improving throughput when sending large messages by avoiding moat checksum(More)
We present detailed measurements of various processing overheads of the TCP/IP and UDP/IP protocol stacks on a DECstation 5000/200 running the Ultrix 4.2a operating system. These overheads include data-touching operations, such as the checksum computation and data movement, which are well known to be major time consumers. In this study, we also considered(More)
Networking software is a growing bottleneck in modern workstations, particularly for high throughput applications such as networked digital video. We measure various components of the UDP/IP protocol stack in a DECstation 5000/200 running Ultrix 4.2a, and find that checksumming and copying dominate the processing time for high throughput applications. This(More)
In this paper, we examine several distributed caching strategies to improve the response time for accessing data over the Internet. By studying several Internet caches and workloads, we derive four basic design principles for large scale distributed caches: (1) minimize the number of hops to locate and access data, (2) do not slow down misses, (3) share(More)
Network software speed is not increasing as rapidly as that of work-station CPUs. The goal of this study is to determine how various components of network software contribute to this bottleneck. In this extended abstract, we summarize the performance of TCPIP and UDPIP networking software for the DECstation 5000/200 workstations connected by an FDDI LAN.
Networking software is a growing bottleneck in modern workstations, particularly for high throughput applications such as networked digital video. We measure various components of the UDP/IP protocol stack in a DECstation 5000/200 running Ultrix 4.2a, and quantify the way in which checksumming and copying dominate the processing time for high throughput(More)
This paper describes the general problems in providing system software support for multimedia applications. We specifically focus on applications which have digital continuous media (DCM) I/O components, and the type of support that must be provided by the network software and operating system. DCM applications have large bandwidth and low delay(More)
distributed applications require the efficient movement of very large objects, from tens to hundreds of megabytes in size. The network architecture incorporates new designs and implementations of operating system I/O software. New methods provide significant performance improvements for transfers among devices and processes and between the two. These(More)