176 citations found. Retrieving documents...
Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the Fifteenth SOSP, pages 40--53, Copper Mountain, CO, December 1995.

 Home/Search   Document Not in Database   Summary   Related Articles   Check  

This paper is cited in the following contexts:

First 50 documents  Next 50

Researching System Administration - Anderson   (Correct)

....in the common case) and 128 MB of memory. All communication is performed with Active Messages (AM) MC96] AM exposes most of the raw performance of Myrinet while providing support for threads, blocking on communication events, and multiple independent endpoints. Other fast message layers [PLC95, vEBBV95, vECGS92] require polling the network interface, which mostly defeats the single node sharing desirable for an I O infrastructure such as River. 4.3.2 Distributed queue performance First, we explore the scaling behavior of the distributed queue. In the first experiment, data is read from n ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the 14th ACM Symposium on Operating Systems Principles, pages 40--53, December 1995.


Damelo! An Explicitly Co-locating Web Cache File System - Ledlie (2000)   (Correct)

.... the LogStructured File System [17] Flushing Damelo s groups metadata in the background and at distinct times comes from [13] and [1] The concept of less than ACID semantics is discussed more fully in [6] Attempting to minimize in memory copying, in particular to a network port, was utilized in [20] s zero copy. Placing large objects toward the outside of the disk platter is discussed more fully in [15] Damelo s bu#er manager and internal interfaces grew out of the Minirel project [5] The speedups possible through in memory metadata were explored by the author and Matthew McCormick in ....

Thorsten von Eicken, Anindya Basu, Vineet Bush, and Werner Vogels. U-net: A userlevel network interface for parallel and distributed computing. In Proceedings of 15th ACM Symposium on Operating Systems Principles, pages 40--53. Association for Computing Machinery SIGOPS, 1995.


Integrating User-Level Networks with SMT - Parker, Davis, Hsieh   (Correct)

....quickly by the processor, and acts as a staging area for outgoing messages. A zero copy message protocol allows messages to be delivered directly to user space without copying. Not all of these ideas are new. For example, previous research has explored the use of user level network interfaces[3,9,11,13,18]. However, this specific combination of features is unique, in that it exposes interrupts directly to user level programs. The important aspect of our architecture lies in its support for user level messaging (for both interprocessor communication and I O) in a general purpose operating system ....

....message arrival notification. Sends, receives, and notifications all make passes through operating system code. Since the operating system code is unlikely to reside in the cache, these system calls result in cache misses. Figure 1: Anatomy of a message for a kernel mode NI User level interfaces[3,9,11,13,18] and zero copy protocols[5,7] significantly reduce the overhead of message sends and receives by eliminating operating system and copying overhead on the message send and receive sides. Notifications still have significant opportunity for optimization, as they remain the performance and ....

[Article contains additional citation context not shown here]

Thorsten von Eicken, et al. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the 15th ACM SOSP, December 1995, pp. 40-53.


Towards Untrusted Device Drivers - Leslie, Heiser (2003)   (2 citations)  (Correct)

.... Other approaches provided user code direct access to network interfaces in order 6 0 10 20 30 40 50 60 70 CPU Transfer size (kB) Disk read CPU Utilisation I2 Linux Mungi Figure 3: CPU utilisation to minimise latency for fine grained communication in high performance clusters [vEBBV95,Dam98] Earlier work with real user level drivers in Mach [GSR93] and Fluke [VM99] experienced significant performance problems, apparently resulting from the IPC costs in those kernels. The Palladium approach of running Linux kernel extensions at an intermediate privilege level [CVP99] could, ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on OS Principles (SOSP), pages 40--53, Copper Mountain, CO, USA, December 1995.


VIA over the CLAN Network - Riddoch, Pope, Mansley (2000)   (2 citations)  (Correct)

....present some initial performance results, and compare it with an existing native hardware implementation. 2 Background 2.1 Virtual Interface Architecture VIA is the result of an effort to standardise academic research into user level networks. The design is most strongly influenced by the U Net[6] project, but a wide variety of other user level networks have been proposed[7, 8, 9] The VIA network model has since been adopted for the Infiniband switched fabric interconnect, which has wide and powerful industry support, and hence is likely to be widely adopted. The VIA Specification[10] ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.


Harnessing User-Level Networking Architectures for.. - Madukkarumukumana, Shah (1998)   (10 citations)  (Correct)

.... overheads associated with underlying legacy transport protocols (e.g. UDP, TCP) used in RPC mechanisms introduce considerable latency over high speed networks such as System Area Networks (SANs) User level networking architectures, such as the Virtual Interface (VI) Architecture [8, 18] U Net [10], and SHRIMP Virtual Memory Mapped Communication (VMMC) 2] that are designed to achieve low latency and high bandwidth in a SAN environment, offer an attractive solution for reducing communication software overheads. Building high level applications, using lowlevel primitives offered by ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels, "U-Net: A User-Level Network Interface for Parallel and Distributed Computing ACM Symposium on Operating System Principles, 1995, pp. 40-53.


VIA over the CLAN Network - Riddoch, Pope, Mansley (2001)   (2 citations)  (Correct)

....present some initial performance results, and compare it with an existing native hardware implementation. 2 Background 2.1 Virtual Interface Architecture VIA is the result of an e ort to standardise academic research into user level networks. The design is most strongly in uenced by the U Net[6] project, but a wide variety of other user level networks have been proposed[7, 8, 9] The VIA network model has since been adopted for the In niband switched fabric interconnect, which has wide and powerful industry support, and hence is likely to be widely adopted. The VIA Speci cation[10] ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A UserLevel Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.


Tesla: A Transparent, Extensible Session-Layer Framework for.. - Salz (2002)   (Correct)

....in an application speci c fashion [8] With Tesla, our goal is to bring some of the power of these systems to commodity operating systems in the context of session layer services. Several other projects have explored user level networking in the context of traditional operating systems. U Net [35] provides direct, user level access to the network interface and an infrastructure to implement user level network stacks. Alpine [13] virtualizes much of the FreeBSD network stack, moving it to the user level, although it is highly platform dependent and intended mostly for debugging networking ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proc. ACM SOSP '95, pages 40-53, December 1995.


Fine-Grain Distributed Shared Memory on Clusters of Workstations - Schoinas (1997)   (3 citations)  (Correct)

....to communicate without kernel intervention and is supported in recent low latency communication hardware [BCF 95,BLA 94, Hor95,OZH 96] Fast access to the hardware by itself is not enough to achieve low latencies. As the experience from message passing multicomputers has shown [vECGS92, PLC95, vEBBV95] traditional messaging interfaces have not been able to realize the hardware performance due to fixed software overheads associated with sending and receiving messages. Therefore, a newer generation of low overhead messaging interfaces attacked the software overheads. Some software architectures ....

....and or memory into the user address space. Thereafter, the application can initiate message operations communicating directly with the device using loads and stores to send and receive messages. Examples of such designs include the Arizona Application Device Channels (ADCs) DPD94] Cornell UNet [vEBBV95] HP Hamlyn [Wi192] Princeton SHRIMP [BDFL96] The result of such recent research has been commercial designs like Myricom Myrinet [BCF 95] Fore 200 [CMSB91] Mitsubishi DART [OZH 96] With the OS removed from the critical path, the memory subsystem emerges as a major hurdle for delivering the ....

[Article contains additional citation context not shown here]

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-net: A user-level network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on Operating System Principles (SOSP), pages 40-53, December 1995.


A Low Overhead Application/Device-driver Interface for.. - Riddoch, Pope (2001)   (Correct)

....on the host[1] Contributions include the cost of making system calls, demultiplexing, protocol stacks and copying of data between addressspaces and buffers. In the last few years a number of projects have addressed these problems by providing a user level interface to the network interface[2, 3]. Applications access the network directly through a virtual memory mapping onto the hardware and or through a region of host memory shared with it. In some cases the network provides reliable data delivery, allowing very simple, low overhead protocols, since there is no need for error detection, ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User- Level Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.


An Analysis of VI Architecture Primitives in Support.. - Begel, Buonadonna.. (2002)   (3 citations)  (Correct)

....server load. However, none of these studies analyze the costs associated in the realization of these protocols or applications upon the VI primitives. There has been extensive work in examining protocol layer costs over other user level communication abstractions for high performance systems [9, 21, 23, 39], and from these we draw much of our methodology. Additionally, we draw upon the lessons of related highperformance network architectures [15, 31, 32] and the benchmark techniques developed by Culler et al. in [13, 14] to analyze our implementations. Specific to the VI Architecture, Speight et. ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-net: a user-level network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on Operating Systems Principles (SOSP), volume 29, pages 303--316, 1995.


Improving Computer Communication Performance by Reducing Memory.. - Ahlgren (1997)   (2 citations)  (Correct)

....single copy design was also used by Partridge and Pink [62] when optimizing the Berkeley UDP implementation in SunOS. Implementing protocols in the application address space is not without problems. Maeda and Bershad [54] Thekkath et al. 73] Edwards and Muir [37] as well as von Eicken et al. [78] report on work in this area. Many researchers working with efficient protocol implementation and network adapters came to the conclusion that the communication performance of workstation class computers is limited by the performance of their primary memory systems, for example, Druschel et al. ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proceedings of the Fifteenth ACM Symposium on Operating Systems Principles, pages 40--53, Copper Mountain Resort, Colorado, USA, December 3--6, 1995. ACM SIGOPS Operating Systems Review, 29(5).


LoPC: Modeling Contention in Parallel Algorithms - Frank, Agarwal, Vernon (1997)   (16 citations)  (Correct)

....Active Messages [33] is general enough to implement more complex communication and synchronization protocols, which we believe makes it a good basis for algorithm analysis in modern parallel systems. A number of existing machines provide efficient support for communication with active messages [2, 9, 14, 20, 24, 28, 32]. Wemake two further assumptions in the LoPC model that appear to result in very little loss in accuracy yet great gain in simplicity. First, we assume that the hardware message buffers at the nodes are infinitely large. Second, we assume that the interconnect is contention free. We model ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the Fifteenth ACM Symposium on Operating Systems Principles, pages 40--53, December 1995.


Dynamic Coscheduling on Workstation Clusters - Sobalvarro, Pakin, Weihl, Chien (1998)   (40 citations)  (Correct)

....and timesharing scheduling (based on kernel scheduler changes) the problem is more difficult for shared workstation clusters in which stock operating systems kernels must be run. With clusters connected by high performance networks that achieve latencies in the range of tens of microseconds [13, 20, 21, 9, 7], scheduling and context switching latency can increase communication latency by several orders of magnitude. For example, under Solaris, CPU quanta vary from 20 ms to 200 ms [10] consequently uncoordinated scheduling can increase best case latencies ( 10 microseconds) by three to four orders of ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on Operating Systems Principles, December 1995. Available from http://www.cs.cornell.edu/Info/Projects/ATM/sosp.ps.


Tripwire: A Synchronisation Primitive for Virtual.. - Riddoch, Pope.. (2000)   (Correct)

....architectures, such as ethernet and ATM, deliver relatively poor performance to real applications, despite impressive increases in raw bandwidth. This has been shown to be largely due to high software overhead. To address this problem a variety of user level network interfaces have been proposed[1, 2, 3, 4, 5, 6]. These have achieved dramatically reduced overhead and latency, and deliver substantial performance benefits to applications. However, these new interfaces are typically unsuitable for use in general purpose local area networks. The interface is often designed to support a particular class of ....

....Channel[3] the Scalable Coherent Interface[12] SHRIMP[6] and BIP[22] All of these have addressed the requirements for high bandwidth, low latency data transfer, providing excellent performance for a particular class of problem, but support for synchronisation is often inflexible. The U Net[1] project was the first to present a userlevel interface to local area networks, using off theshelf communications hardware. U Net heavily influenced the Virtual Interface Architecture[24] which provides a communication model based on asynchronous message passing through work queues. Completion ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.


Multigrain Shared Memory - Yeung, Kubiatowicz, Agarwal (2000)   (Correct)

.... across a wide range of applications [2] While communication interfaces for commodity workstations have made impressive improvements, the best reported inter workstation latency numbers are still an order of magnitude higher than on machines that have tightly coupled special purpose interfaces [3]. The high cost of communication on commodity systems prevents them from supporting applications with intensive communication requirements. Traditional architectures for large scale shared memory machines have not satisfactorily addressed the tension between providing efficient communication ....

....our results do not reflect the impact of contention in the inter node network and at the inter node network interfaces. Second, by using a latency of 50 sec, our results only reflect the performance of DSMPs with aggressive networks. While a 50 sec one way messaging latency is achievable (e.g. [3] reports a latency of 33 sec over ATM) such low latency is possible only on high performance networks. In this section, we explore the impact that higher communication latency, due to either contention or a slower network, has on our results. Our study exploits the hardware timers used to set ....

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the 15th ACM Symposium on Operating Systems Principles, Copper Mountain, Colorado, December 1995.


A Software Approach for Readout and Data Acquisition in .. - Antchev, Cano..   (Correct)

....of highly efficient programs. E. Zero Copy Streams Figure 6: Components encapsulate various different tasks, but expose the same clean and narrow interface. The implementation of the I O streams aim at achieving good performance. Therefore they rely on zero copy buffers wherever possible [16] [17], 18] In general, driver interfaces Operating system OS provided buffers OS provided buffers Operating system Zero Copy Buffer Stream Component Copy Buffer Stream Component UDP RingBuf Zcopy UDP Zcopy RingBuf 1 Control Flow Data Flow FU Output Stream RUM Input Stream ....

T. von Eicken, A. Basu, V. Buuch and W. Vogels, U-Net: A User-Level Network Interface for Parallel and Distributed Computing, in Proceedings of the Fifteenth ACM Symposium on Operating System Principles, pages 40-53, ACM Press, December 3-6, Copper Mountain Resort, Colorado, USA, 1995.


Implicit Coscheduling: Coordinated Scheduling with Implicit.. - Arpaci-Dusseau (1998)   (5 citations)  (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In Proceedings of the Fifteenth SOSP, pages 40--53, Copper Mountain, CO, December 1995.


User-level Device Drivers: Achieved Performance - Leslie, Chubb, Fitzroy-Dale, .. (2005)   (1 citation)  (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In 15th SOSP, pages 40--53, Copper Mountain, CO, USA, Dec 1995.


Analyzing NIC Overheads in Network-Intensive Workloads - Binkert, Hsu, Saidi.. (2004)   (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proc. Fifteenth ACM Symp. on Operating System Principles (SOSP), pages 40--53, 1995.


Market-based Cluster Resource Management - Chun (2001)   (2 citations)  (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-net: A user-level network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on Operating Systems Principles, pages 40-53, Copper Mountain, CO, December 1995.


Encapsulated User-Level Device Drivers in the Mungi Operating.. - Nicholas   (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A userlevel network interface for parallel and distributed computing. In Proc. 15th SOSP, pages 40--53, Copper Mountain, CO, USA, Dec 1995.


Achieving Gigabit Performance on Programmable Ethernet Network.. - Wadge (2001)   (2 citations)  (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A userlevel network interface for parallel and distributed computing. In Proceedings of the 15th ACM Symposium on Operating Systems Principles, December 1995


Distributed Computing with the CLAN Network - Riddoch, Mansley, Pope (2002)   (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.


Tripwire: A Synchronisation Primitive for Virtual.. - Riddoch, Pope.. (2000)   (Correct)

No context found.

Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A User-Level Network Interface for Parallel and Distributed Computing. In 15th ACM Symposium on Operating Systems Principles, December 1995.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC