Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications

Size: px
Start display at page:

Download "Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications"

Transcription

1 Year 2Activities report for the NSF project EIN Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Date: July 15, 2005 PI: Malathi Veeraraghavan, Please reiterate the goals and objectives of your efforts, and summarize the research and education activities you have engaged in that aim to achieve these objectives. Include experiments you have conducted, the simulations you have run, the collecting you have done, the observations you have made, the materials you have developed, and major presentations you have made about your efforts. In a later section you will list more formally any publications and other specific products (database, collections, software, inventions, etc.) that have resulted. The goal of this project is to develop the infrastructure and networking technologies to support a broad class of escience projects and specifically the Terascale Supernova Initiative. Our objectives are to design and deploy a high-performance, experimental optical network infrastructure and to test application/ middleware/transport protocol software, developed specifically for escience projects, on this network. Our two target applications are file transfers and remote visualization. The objective of our research activities is to determine whether call-by-call dynamic sharing of bandwidth resources can be used for data applications. Specifically, can such sharing reduce the costs of highspeed circuits, thus increasing their availability for escience applications such as the TSI project. In addition, we are designing and implementing transport protocols that handle the mismatch between the characteristics of multi-tasking general-purpose end hosts with off-the-shelf disks with the characteristics of circuits. To achieve high circuit utilization, the sending end host should be programmed to send data at a monotonic rate equal to that of the circuit, and the receiving end host should be programmed to move data from kernel buffers to the disk at that same rate. The objective of the educational activities is to prepare a new generation of engineers who are knowledgeable about the new technological advents in the general area of optical networks and specifically in how to leverage the high-speed capabilities of these networks to architect and design new applications. In addition, our students have the opportunity to acquire inter-disciplinary skills spanning operating systems, distributed systems, parallel computing and networking.

2 Below is a summary list of all our activities in this project to date (Aug July 2005), which includes experiments we have conducted, and some observations we have made (more details on our observations are listed in the findings attachment): CHEETAH wide-area network (primarily work of postdoctoral fellow, Xuan Zheng, with support from graduate student, Xiangfei Zhu): We completed design of the network, negotiated with vendors, tested equipment, purchased and installed equipment, and finally interconnected network nodes with circuits. Activities included testing switches, ordering equipment (including SONET-based MSPPs, high-end computers, Ethernet switches, Netscreen security devices, GbE network interface cards, console servers and power distribution units (for remote control), fiber cables and other minor associated pieces of equipment), negotiating with four service providers to purchase various wide-area optical circuits, finalizing order agreements and purchasing these circuits, negotiating and purchasing collocation services to locate our switch equipment in two PoPs (RTP area in NC and Atlanta). We now have a running highspeed circuit-switched network in place as shown in Fig. 1. [We request the reader to view Fig. Figure 1 CHEETAH network as of July 7, on a computer screen rather than on a black-and-white printout for two reasons: (a) color and (b) ability to enlarge the figure.] We have deployed three Sycamore SN16000 switches as shown in Fig. 1, using collocation agreements with MCNC and SLR/SOX (space and power).

3 Each SN16000 has multiple GbE ports (the ORNL switch additionally has a 10GbE port), and multiple OC192 ports. These nodes run GMPLS signaling and routing protocol software, version 7.6. We also purchased and located computers (wukong in MCNC) and zelda1, zelda2, zelda3 (at SOX), and zelda4 and zelda5 at ORNL. These computers run Redhat Linux version 2.6 and have two GbE NICs each, one connected to the Internet (primary) and the second connected to the CHEETAH network. We purchased Internet access as part of our collocation service agreements from NCNI (MCNC) and SLR/SOX. The wide-area OC192 circuits were purchased from (1) MCNC (from MCNC in the Research Triangle Park to NLR PoP in Raleigh) via Cisco MSTPs, (2) NLR (for the Raleigh-Atlanta segment) via Cisco 15808s, (3) SLR (for intra-atlanta segment from the NLR PoP to the SoX PoP) via Movaz Ray Express equipment, and (4) ORNL (for the SOX-ORNL segment) via Ciena Corestream equipment. Non-recurring costs include the purchase of OC192 transponders for all four segments. The first three are functional and in use; only the last segment still needs to be activated. In its place we are using 1Gb/s MPLS tunnels as circuits (up to 3 Gbps) on an existing OC192 circuit extending between two Juniper M320 routers. The circuits from NCSU (Centaur Lab), where our co-pi John Blondin s compute cluster, orbitty, is located, to MCNC, are realized using five 1Gbps VLANs. A 10GbE circuit was already in place between NCSU s Centaur Lab and NCSU s CMDF building, and between the CMDF building and MCNC. These five VLANs were provisioned through cooperation between NCSU and MCNC. The Foundry FESX448 and the two Cisco 7600 Ethernet switches shown in Fig. 1 needed to be manually configured for the VLANs. We have successfully tested communication between various pairs of computers across various segments of our network, e.g., orbitty to wukong, orbitty to zelda1/2/3, orbitty to zelda 4/5. Application support for TSI (primarily work of graduate student, Xiuduan Fang): We downloaded and tested PVFSv2 (Parallel Virtual File System) and GridFTP on a local cluster in our CS department. We found several problems in adapting these to the CHEETAH network and modified the source code accordingly. The two main problems were (1) PVFSv2 tools did not actually support the -s option that was available in the API to set the stripe size, and (2) there was no way to specify the stripe order and control the physical location of stripes. In the CHEETAH concept, we set up dedicated circuits host-to-host. Thus if there are two clusters orbitty and zelda, and circuits are set up between orbitty1 and zelda1, orbitty2 and zelda2 and so on, then we wanted to control how PVFS striped a file on zelda (located in ORNL) so that the stripe on zelda1 was transferred to orbitty1, and so on. Using strace, Xiuduan determined

4 what parts of the PVFS code needed modification and did so. The GridFTP striped transfer mode is used to enable multiple nodes to work together on a single file and act as a single GridFTP server. It relies on an underlying parallel file system (e.g., PVFS2) to allow all nodes to see the same file system. It uses third-party transfer, in which a controller running on an external end host initiates a transfer from one cluster to another cluster. The GridFTP protocol extends the FTP PASV and PORT commands to SPAS and SPOR commands respectively, setting the receiving cluster to operate in passive mode, and obtaining the port numbers on which this cluster s servers are listening, and providing these port numbers to the sending cluster allowing the latter s servers to initiate data connections to the receiving servers. In the current GridFTP code, the process of assigning the port numbers returned in response to the SPAS command to the sending cluster is non-deterministic and in each run a different pair of hosts could be required to connect. This makes the current implementation unsuitable for CHEE- TAH unless the SPAS and SPOR code is modified. We have currently suspended this activity for two reasons. First, our TSI project co-pi, John Blondin, clarified that while the TSI simulations generate TB-sized datasets, these datasets are created in the form of several files, each on the order of gigabytes. Since each host on his orbitty cluster has three 146GB disks, different files can be stored on different compute nodes of orbitty. In other words, the TSI project does not necessarily need to store single large files striped across the local disks of several computers, which is the primary reason for needing PVFS. In general, even if a file can fit into the disk of one host, there are advantages to striping it across the disks of multiple end hosts and then use striping to achieve parallelism in the transfer. However, in the TSI project since there are multiple GB-sized files, we exploit parallelism across the multiple files generated in a simulation run rather than within a single file transfer. Second, we extended the CHEETAH solution to allow the concept of a cluster-to-cluster dedicated circuit. Since cluster computers nowadays increasingly use GbE switches, we propose to connect one port of a cluster GbE switch to the CHEETAH network and other ports to the second NICs on the cluster computers. By plugging the switch port into the CHEETAH Sycamore switch, we can create Ethernet switch-to-ethernet switch connections. This would allow Ethernet frames destined to any single cluster node at one end from any node on the far-end cluster to reach successfully without requiring pairwise dedicated circuits. We are now experimenting with performance results from FTP over tuned TCP and FRTP, as well as bbcp and SecureFTP applications. After a dedicated circuit is set up, there is almost no possibility of packet loss with the use of TCP because the only buffer involved is the receiver s buffer, and TCP s window based flow con-

5 trol adequately warns the sender. With GB sized files, the loss of throughput in the initial Slow Start is marginal. General-use applications (primarily work of graduate student, Xiuduan Fang): We used the Common Gateway Interface (CGI) mechanism to integrate CHEETAH software into the web browsing application to trigger the automatic set up of a circuit and the use of it for downloading the requested web page. This was demonstrated at SuperComputing Signaling software (primarily work of graduate student, Xiangfei Zhu): We designed and implemented an RSVP-TE client for Linux hosts starting with RSVP-TE Kom/Dragon code. We tested interoperability of this client with Sycamore s SN16000 RSVP-TE implementation. This test was a success. We found a few bugs in our code and fixed them. The availability of RSVP-TE decoding in the sniffer tool, Ethereal, greatly helped us in this task. Sycamore s implementation of the GMPLS protocol suite is quite good. However, a significant part of our activities were used to identify the boundaries of their implementation and to work within these boundaries. For example, given there is no standard yet for signaling for heterogeneous connections, and in the CHEETAH network we necessarily have heterogeneous connections (Ethernet to hosts and SONET between switches), we needed GMPLS support for heterogeneous connections. Sycamore developers were highly cooperative and provided us a proprietary solution for this problem, with a commitment to evolve this to a standards based solution once that is defined. Other features we found lacking in the Sycamore implementation are as follows: we cannot define multiple TE links per port with corresponding capacity levels (split port capacity among TE links), specify a subnet mask for remote IP addresses when configuring TE-links, and cannot specify multiple IP addresses per TE-link. This is because the Sycamore software was developed for a transit node, not for an ingress/egress node, i.e., nodes on which the circuit terminates. These limitations do pose constraints, but we have designed our network and software within these constraints. We also integrated the Dynamic TL1 library purchased from Monfox with RSVP-TE software to control the Cisco externally. We demonstrated the setup/release of circuits through the at SuperComputing We borrowed high-speed GbE cards from MCNC/Cisco and later purchased an ML GbE card for our laboratory testbed. Transport protocol (primarily work of graduate student, Anant P. Mudambi): The main problem we tackled this year was to have the sending end of the transport protocol implementation maintain a constant sending rate to match the rate of the circuit. This is important to achieve high circuit utilization. We started off with the FRTP1.0 (Fixed Rate Transport Protocol) code,

6 which was modified from a SABUL implementation. But this implementation used a busywait mechanism at the sender to maintain the required inter-packet gap. This not only made it a CPU-intensive implementation, but also made the sender s rate very sensitive to any interruptions (e.g., if another process becomes scheduled to run or the sending process blocks in some system call like while reading from disk). We tried a second approach to this problem of maintaining a constant sending rate by using a larger time frame; in other words, have the sender send a fixed number of packets every T time units. We tried this by modifying the FRTP1.0 code to send a fixed number of packets and then sleep for some time. The packets are then sent in a burst. To reduce burstiness, we would like T to be as small as possible. In Linux 2.4, we found that the minimum value for T is 10ms, but none of the system calls related to sleep were accurate at such small time intervals. What we needed was a way for the user-space sending process to get a reliable interrupt every T time units. Linux offers a way for a program to set a timer and have the OS send a signal to the process whenever the timer expires. We used this mechanism to implement a more accurate way for the sending process to be invoked periodically. This worked fairly well, except the burst sending of packets caused a problem if the SONET wide-area circuit rate was lower than the end host NIC rate of 1Gb/s. In the CHEETAH solution, we proposed such a circuit to handle the issue that receive disk rates are more commonly in the 400Mbps range, and hence wanted to limit the bandwidth wasted in the wide-area circuit. Luckily, both the Cisco MSPP and Sycamore s SN16000 Ethernet cards support the GbE PAUSE feature if the GbE port is mapped to a lower rate SONET circuit. This PAUSE signal, sent by the GbE card in the MSPP stops the sender s Ethernet driver from sending frames. While this solution appears to be perhaps the best choice for a user-space based FRTP implementation, it is not very accurate in maintaining a constant rate if the sender is multi-tasking. If the file transfer sending process is not running when the timer signal expires, the process will not be invoked immediately and hence it will miss being strictly monotonic. Another problem we identified with a user-space implementation is that windowbased flow control is difficult to implement because the size of the UDP buffer, held in the kernel, is the one that needed reporting back to the sender and this size is difficult to obtain in time before the receive window overflows. Therefore, we spent the second half of the year on a kernel based implementation of FRTP using Web100. The goal was to reuse window control implemented as part of the TCP kernel implementation. To remove TCP s congestion control (given this is detrimental when using a circuit), we downloaded Net100/Web100, which provides a way to tune the AIMD parameters of TCP. Web100 does not provide a way to set the

7 AI and MD parameters to 0, which is what we require to be able to maintain a steady congestion window. We added code to the Linux kernel TCP stack to be able to do this. The shift to a kernel-based implementation reusing TCP and Web100 code provides another important advantage: TCP s self-clocking provides a way to maintain a steady sending rate. Once a congestion window s worth of data is sent into the network, the sender sends the next packet only when a returning ACK slides the window to the right. The ACKs come back at the bottleneck link rate, which in our case, is the circuit rate, and so the sending rate equals the circuit rate. We thus have two new implementations of FRTP, one user-space UDP based, and a second kernel-space TCP based. We are currently measuring the performance of these implementations and integrating them with applications. Routing decision software (primarily work of graduate student, Zhanxiang Huang): The goal of the routing decision module is to decide whether a transfer should attempt the setup of a circuit or directly use the Internet path. The models for computing the cost of circuit setup and the TCP/IP connectionless path from the delay and utilization aspect have already been described in the CHEETAH Opticomm 2003 paper and Dr. Xuan Zheng s dissertation. This year we focussed on designing and implementing the routing decision module. We defined the requirements for the routing decision module as follows. It should handle multiple queries from application clients simultaneously. It should maintain a cache of frequently queried endpoints to reduce query overhead. The design is based on multithreading and synchronization. We specified the interfaces between sub-modules, and identified the parameters that need to be measured and also selected appropriate measurement tools after experimenting with several tools, such as pchar, pathload, pipechar, iperf, etc. We designed the cache structure, and implemented the delay model for TCP and circuit setup/transfer. The parameters we measure include round trip time (RTT) on both the connectionless and circuit path, the packet loss rate on the connectionless path, and TCP client side maximum window size. Analysis of remote visualization application traffic (primarily work of graduate student, Zhanxiang Huang): The goal of monitoring the traffic of the TSI visualization application, Ensight, is to help us determine what type of circuit should be set up for remote visualization. The monitoring environment is as follows. Professor Blondin runs Ensight client on his office desktop to connect to the Ensight server running on a powerful remote cluster called orbitty. We placed a fiber splitter between the orbitty cluster and its outgoing switch and the splitter forwarded all the traffic from and to orbitty to another host named cheetah.ncren.net, on which we captured and analyzed the traffic. We used tcpdump as the capturing tool, and with its fil-

8 tering function, we filtered out other traffic irrelevant to the Ensight application. Since we already knew that the Ensight server sends one geometry structure data at a time, the main goal of the analysis was to find out the size of a single geometry structure. We first used Ethereal s post-processing tool to separate the server-to-client and client-to-server traffic, and then used the statistical tool provided in Ethereal to determine the peak burst rate. However, Ethereal could not directly provide us the size of a typical geometry structure. Thus we had to write our own post-processing tool to analyze it. Our key assumption was that one geometry structure is sent by the server application to the kernel TCP buffer at once, and expected this to be reflected in the traffic as a sequence of contiguous full-mtu length packets. Further, we expected control messages to be sent in smaller packets interleaved among the sequence of packets carrying the geometry data. Thus the first effort of our post-processing program was to find the sequence of contiguous full-mtu length packets. But it turned out that our assumption that the Ensight server application sends a whole geometry structure to the TCP buffer was wrong. In fact it breaks this up into smaller blocks. Using this assumption, we induced there should be at most one packet whose size is less than the full MTU length in between blocks, which is highly consistent with our observation of the captured trace file. Furthermore, we used the tool mergecap, to synchronize the server-to-client and client-to-server traces and then plotted the TCP congestion window size against time to check the window size when the smaller packets occur with the tool tcptrace and xplot. This confirmed that the smaller packets are not caused by the change of TCP congestion window size. Thus we modified our own post-processing tool to find the contiguous sequence of full-mtu length packets with at most one smaller sized packet in between. The analysis results showed us that the maximum geometry structure data we observed is around 3.216MB and the time it took to transfer this geometry is around 0.33 second. The interval in between sending geometries from server to client depends on the frequency of the user request from the client. Thus our preliminary conclusion is that remote visualization does not require a high-bandwidth circuit, and if we can predict the user request model, we can decide when to set up and when to tear down a circuit. Project management: We coordinated activities between all four participating organizations. We supported the CUNY team in their development of VLSR (with design and OSPF testing), and security solution for CHEETAH. We worked with the ORNL and NCSU teams to facilitate the TSI application support for file transfers and remote visualization. Education activities include the following. Xiangfei Zhu obtained his Masters degree in May He is continuing on with support of this project for a Ph.D. Two undergraduate students, Andrew Love and

9 Yan Peng (Roger) Guo, did their undergraduate senior thesis projects in our laboratory. Students have taken full advantage of the opportunities provided in realizing this wide-area testbed and learning the use of all the equipment we have procured. Significant learning about network security devices was an important aspect of our educational growth this year. Materials: The materials we developed include papers and software. The publications are listed in the Products part of this report. We published two journal papers, one conference paper and one workshop paper. Software developed in the project along with presentations are being posted on our project web site: cheetah.cs.virginia.edu/. Presentations (the first by Xuan Zheng, and the rest by Malathi Veeraraghavan): Title of Presentation Meeting Place Date End-To-End Provisioned Optical Network Testbed for Large-Scale escience Applications CHEETAH CHEETAH Joint Engineering Team Review of Optical Networking Testbeds, NSF, Washington, DC NSF Sponsored Planning Workshop on The Future of Optical Communications: Understanding the Choices Virginia Tech High-Performance Computing Conference Blacksburg Washington, DC Santa Barbara May 26-27, 2005 April 19, 2005 April 12 & 13, 2005 Enabling a connection-oriented internet Georgia Tech Atlanta March 30, 2005 Enabling a connection-oriented internet Duke University Durham, Feb. 11, 2005 NC CHEETAH Exhibition SuperComputing 2005 Pittsburgh Nov. 6-12, 2004 Immediate-request vs. scheduled calls, Short-duration vs. long-duration calls Served as panelist MCNC Workshop on GMPLS Control Plane Optical Networks and Grid Computing Panel, Broadnets, 2004 Pittsburgh November 12, San Jose Oct , 2004

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Year 1 Activities report for the NSF project EIN-0335190 Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Date: July 29, 2004 (this is

More information

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Year 3 Activities report for the NSF project EIN-0335190 Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Date: July 21, 2006 PI: Malathi

More information

A TRANSPORT PROTOCOL FOR DEDICATED END-TO-END CIRCUITS

A TRANSPORT PROTOCOL FOR DEDICATED END-TO-END CIRCUITS A TRANSPORT PROTOCOL FOR DEDICATED END-TO-END CIRCUITS MS Thesis Final Examination Anant P. Mudambi Computer Engineering University of Virginia December 6, 2005 Outline Motivation CHEETAH Background UDP-based

More information

Improving web performance through new networking technologies

Improving web performance through new networking technologies Improving web performance through new networking technologies Xiuduan Fang, Xuan Zheng, Malathi Veeraraghavan University of Virginia {xf4c, xz3y, mv}@virginia.edu Abstract New connection-oriented networking

More information

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications

Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Year 4 Activities report for the NSF project EIN-0335190 Title: Collaborative research: End-to-End Provisioned Optical Network Testbed for Large-Scale escience Applications Date: July 21, 2007 PI: Malathi

More information

CTCP (Circuit TCP) v1.0

CTCP (Circuit TCP) v1.0 Outline C (Circuit ) v1.0 Helali Bhuiyan and Malathi Veeraraghavan {helali,mv5g}@virginia.edu Purpose of C Requirements to run C code C Components C Operation C details C Usage Usage of C across an Internet

More information

Report on Transport Protocols over Mismatched-rate Layer-1 Circuits with 802.3x Flow Control

Report on Transport Protocols over Mismatched-rate Layer-1 Circuits with 802.3x Flow Control Report on Transport Protocols over Mismatched-rate Layer-1 Circuits with 82.3x Flow Control Helali Bhuiyan, Mark McGinley, Tao Li, Malathi Veeraraghavan University of Virginia Email: {helali, mem5qf, taoli,

More information

Final Technical Report

Final Technical Report DOE Award Number: DE-FG02-04ER25640 Name of recipient: University of Virginia Final Technical Report Project title: Enabling Supernova Computations by Integrated Transport and Provisioning Methods Optimized

More information

Addressing and secure control-plane network design in GMPLS networks

Addressing and secure control-plane network design in GMPLS networks Addressing and secure control-plane network design in GMPLS networks Abstract 1 : Malathi Veeraraghavan, Xuan Zheng, Xiangfei Zhu {malathi, xuan, xz4p}@virginia.edu April 7, 2006 This document describes

More information

Implementation of a GMPLS-based Network with End Host Initiated Signaling

Implementation of a GMPLS-based Network with End Host Initiated Signaling Implementation of a -based Network with End Host Initiated Signaling X. Zhu, X. Zheng, M. Veeraraghavan Z. Li, Q. Song, I. Habib N. S. V. Rao University of Virginia City University of New York Oak Ridge

More information

Annual Report:

Annual Report: Annual Report: 0087487 Annual Report for Period:09/2001-09/2002 Submitted on: 07/06/2004 Principal Investigator: Veeraraghavan, Malathi. Award ID: 0087487 Organization: Polytechnic Univ of NY Title: Towards

More information

A discussion of goals and control-plane implications for our HOPI experiments

A discussion of goals and control-plane implications for our HOPI experiments A discussion of goals and control-plane implications for our HOPI experiments Malathi Veeraraghavan, Xiangfei Zhu, Tao Li May 5, 2007 [Requesting feedback/comments to mvee@virginia.edu] In Feb. 2007, we

More information

Networks & protocols research in Grid5000 DAS3

Networks & protocols research in Grid5000 DAS3 1 Grid 5000 Networks & protocols research in Grid5000 DAS3 Date Pascale Vicat-Blanc Primet Senior Researcher at INRIA Leader of the RESO team LIP Laboratory UMR CNRS-INRIA-ENS-UCBL Ecole Normale Supérieure

More information

Steven Carter. Network Lead, NCCS Oak Ridge National Laboratory OAK RIDGE NATIONAL LABORATORY U. S. DEPARTMENT OF ENERGY 1

Steven Carter. Network Lead, NCCS Oak Ridge National Laboratory OAK RIDGE NATIONAL LABORATORY U. S. DEPARTMENT OF ENERGY 1 Networking the National Leadership Computing Facility Steven Carter Network Lead, NCCS Oak Ridge National Laboratory scarter@ornl.gov 1 Outline Introduction NCCS Network Infrastructure Cray Architecture

More information

UltraScience Net Update: Network Research Experiments

UltraScience Net Update: Network Research Experiments UltraScience Net Update: Network Research Experiments Nagi Rao, Bill Wing, Susan Hicks, Paul Newman, Steven Carter Oak Ridge National Laboratory raons@ornl.gov https://www.csm.ornl.gov/ultranet February

More information

Chapter 4. Routers with Tiny Buffers: Experiments. 4.1 Testbed experiments Setup

Chapter 4. Routers with Tiny Buffers: Experiments. 4.1 Testbed experiments Setup Chapter 4 Routers with Tiny Buffers: Experiments This chapter describes two sets of experiments with tiny buffers in networks: one in a testbed and the other in a real network over the Internet2 1 backbone.

More information

High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK

High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK [r.tasker@dl.ac.uk] DataTAG is a project sponsored by the European Commission - EU Grant IST-2001-32459

More information

A Deployable Framework for Providing Better Than Best-Effort Quality of Service for Traffic Flows

A Deployable Framework for Providing Better Than Best-Effort Quality of Service for Traffic Flows A Deployable Framework for Providing Better Than Best-Effort Quality of Service for Traffic Flows Proposal Presentation Raheem A. Beyah July 10, 2002 Communications Systems Center Presentation Outline

More information

The Abilene Observatory and Measurement Opportunities

The Abilene Observatory and Measurement Opportunities The Abilene Observatory and Measurement Opportunities Rick Summerhill, Director Network Research, Architecture, and Technology, Internet2 CONMI Workshop / PAM 2005 Boston, MA March 30, 05 Outline History

More information

Path-based networking: From POTS to SDN. Outline

Path-based networking: From POTS to SDN. Outline Path-based networking: From POTS to SDN Malathi Veeraraghavan University of Virginia April 28, 2014 Talk at CUHK, Dept. of IE Thanks to the US DOE ASCR for grant DE-SC0007341, and NSF grants CNS-1116081,

More information

Experiences with Dynamic Circuit Creation in a Regional Network Testbed

Experiences with Dynamic Circuit Creation in a Regional Network Testbed This paper was presented as part of the High-Speed Networks 2011 (HSN 2011) Workshop at IEEE INFOCOM 2011 Experiences with Dynamic Circuit Creation in a Regional Network Testbed Pragatheeswaran Angu and

More information

Send documentation comments to You must enable FCIP before attempting to configure it on the switch.

Send documentation comments to You must enable FCIP before attempting to configure it on the switch. CHAPTER 9 (Fibre Channel over IP) is an IETF standards based protocol for connecting Fibre Channel SANs over IP based networks. encapsulates the FCP frames in a TCP/IP packet which is then sent across

More information

On the use of connection-oriented networks to support Grid computing

On the use of connection-oriented networks to support Grid computing On the use of connection-oriented networks to support Grid computing Malathi Veeraraghavan, Xuan Zheng, Zhanxiang Huang University of Virginia {malathi, xuan, zh4c}@virginia.edu Abstract -- The vision

More information

Zhengyang Liu University of Virginia. Oct 29, 2012

Zhengyang Liu University of Virginia. Oct 29, 2012 SDCI Net: Collaborative Research: An integrated study of datacenter networking and 100 GigE wide-area networking in support of distributed scientific computing Zhengyang Liu University of Virginia Oct

More information

CS 5520/ECE 5590NA: Network Architecture I Spring Lecture 13: UDP and TCP

CS 5520/ECE 5590NA: Network Architecture I Spring Lecture 13: UDP and TCP CS 5520/ECE 5590NA: Network Architecture I Spring 2008 Lecture 13: UDP and TCP Most recent lectures discussed mechanisms to make better use of the IP address space, Internet control messages, and layering

More information

Chapter III. congestion situation in Highspeed Networks

Chapter III. congestion situation in Highspeed Networks Chapter III Proposed model for improving the congestion situation in Highspeed Networks TCP has been the most used transport protocol for the Internet for over two decades. The scale of the Internet and

More information

GUIDE. Optimal Network Designs with Cohesity

GUIDE. Optimal Network Designs with Cohesity Optimal Network Designs with Cohesity TABLE OF CONTENTS Introduction...3 Key Concepts...4 Five Common Configurations...5 3.1 Simple Topology...5 3.2 Standard Topology...6 3.3 Layered Topology...7 3.4 Cisco

More information

Outline Computer Networking. TCP slow start. TCP modeling. TCP details AIMD. Congestion Avoidance. Lecture 18 TCP Performance Peter Steenkiste

Outline Computer Networking. TCP slow start. TCP modeling. TCP details AIMD. Congestion Avoidance. Lecture 18 TCP Performance Peter Steenkiste Outline 15-441 Computer Networking Lecture 18 TCP Performance Peter Steenkiste Fall 2010 www.cs.cmu.edu/~prs/15-441-f10 TCP congestion avoidance TCP slow start TCP modeling TCP details 2 AIMD Distributed,

More information

Flow-start: Faster and Less Overshoot with Paced Chirping

Flow-start: Faster and Less Overshoot with Paced Chirping Flow-start: Faster and Less Overshoot with Paced Chirping Joakim Misund, Simula and Uni Oslo Bob Briscoe, Independent IRTF ICCRG, Jul 2018 The Slow-Start

More information

Impact of Short-lived TCP Flows on TCP Link Utilization over 10Gbps High-speed Networks

Impact of Short-lived TCP Flows on TCP Link Utilization over 10Gbps High-speed Networks Impact of Short-lived TCP Flows on TCP Link Utilization over Gbps High-speed Networks Lin Xue, Chui-hui Chiu, and Seung-Jong Park Department of Computer Science, Center for Computation & Technology, Louisiana

More information

The Measurement Manager Modular End-to-End Measurement Services

The Measurement Manager Modular End-to-End Measurement Services The Measurement Manager Modular End-to-End Measurement Services Ph.D. Research Proposal Department of Electrical and Computer Engineering University of Maryland, College Park, MD Pavlos Papageorgiou pavlos@eng.umd.edu

More information

The Case for Informed Transport Protocols

The Case for Informed Transport Protocols The Case for Informed Transport Protocols Stefan Savage Neal Cardwell Tom Anderson University of Washington Our position Wide-area network performance: is important is limited by inefficient congestion

More information

Enabling New Applications with Optical Circuit- Switched Networks

Enabling New Applications with Optical Circuit- Switched Networks Enabling New Applications with Optical Circuit- Switched Networks A Dissertation Presented to the Faculty of the School of Engineering and Applied Science University of Virginia In Partial Fulfillment

More information

The New Internet2 Network

The New Internet2 Network The New Internet2 Network Director Network Research, Architecture, and Technologies Internet2 GLIF Meeting 11 September 2006 Tokyo, Japan Agenda Basic Ideas Design Ideas Topology Optical nodes Control

More information

IX: A Protected Dataplane Operating System for High Throughput and Low Latency

IX: A Protected Dataplane Operating System for High Throughput and Low Latency IX: A Protected Dataplane Operating System for High Throughput and Low Latency Belay, A. et al. Proc. of the 11th USENIX Symp. on OSDI, pp. 49-65, 2014. Reviewed by Chun-Yu and Xinghao Li Summary In this

More information

Networking for Data Acquisition Systems. Fabrice Le Goff - 14/02/ ISOTDAQ

Networking for Data Acquisition Systems. Fabrice Le Goff - 14/02/ ISOTDAQ Networking for Data Acquisition Systems Fabrice Le Goff - 14/02/2018 - ISOTDAQ Outline Generalities The OSI Model Ethernet and Local Area Networks IP and Routing TCP, UDP and Transport Efficiency Networking

More information

Student ID: CS457: Computer Networking Date: 3/20/2007 Name:

Student ID: CS457: Computer Networking Date: 3/20/2007 Name: CS457: Computer Networking Date: 3/20/2007 Name: Instructions: 1. Be sure that you have 9 questions 2. Be sure your answers are legible. 3. Write your Student ID at the top of every page 4. This is a closed

More information

Application Layer Switching: A Deployable Technique for Providing Quality of Service

Application Layer Switching: A Deployable Technique for Providing Quality of Service Application Layer Switching: A Deployable Technique for Providing Quality of Service Raheem Beyah Communications Systems Center School of Electrical and Computer Engineering Georgia Institute of Technology

More information

perfsonar Deployment on ESnet

perfsonar Deployment on ESnet perfsonar Deployment on ESnet Brian Tierney ESnet ISMA 2011 AIMS-3 Workshop on Active Internet Measurements Feb 9, 2011 Why does the Network seem so slow? Where are common problems? Source Campus Congested

More information

Networking and TCP/IP. John Kalbach November 8, 2004

Networking and TCP/IP. John Kalbach November 8, 2004 Networking and TCP/IP John Kalbach kalbach@psu.edu November 8, 2004 Topology Bus Ethernet, Wireless Star ATM, Switched Ethernet, Gigabit Ethernet Ring Token Ring, FDDI DS Digital Signals Speeds DS 0 64Kbs

More information

ECE 650 Systems Programming & Engineering. Spring 2018

ECE 650 Systems Programming & Engineering. Spring 2018 ECE 650 Systems Programming & Engineering Spring 2018 Networking Transport Layer Tyler Bletsch Duke University Slides are adapted from Brian Rogers (Duke) TCP/IP Model 2 Transport Layer Problem solved:

More information

CS 428/528 Computer Networks Lecture 01. Yan Wang

CS 428/528 Computer Networks Lecture 01. Yan Wang 1 CS 428/528 Computer Lecture 01 Yan Wang 2 Motivation: Why bother? Explosive growth of networks 1989, 100,000 hosts on the Internet Distributed Applications and Systems E-mail, WWW, multimedia, distributed

More information

Experimental Results On Data Transfers Over Dedicated Channels

Experimental Results On Data Transfers Over Dedicated Channels Experimental Results On Data Transfers Over Dedicated Channels Nageswara S. V. Rao, Qishi Wu, Steven M. Carter, William R. Wing Computer Science and Mathematics Division Oak Ridge National Laboratory Oak

More information

Avaya ExpertNet Lite Assessment Tool

Avaya ExpertNet Lite Assessment Tool IP Telephony Contact Centers Mobility Services WHITE PAPER Avaya ExpertNet Lite Assessment Tool April 2005 avaya.com Table of Contents Overview... 1 Network Impact... 2 Network Paths... 2 Path Generation...

More information

TCP and Congestion Control (Day 1) Yoshifumi Nishida Sony Computer Science Labs, Inc. Today's Lecture

TCP and Congestion Control (Day 1) Yoshifumi Nishida Sony Computer Science Labs, Inc. Today's Lecture TCP and Congestion Control (Day 1) Yoshifumi Nishida nishida@csl.sony.co.jp Sony Computer Science Labs, Inc 1 Today's Lecture Part1: TCP concept Part2: TCP detailed mechanisms Part3: Tools for TCP 2 1

More information

Uncompressed HD Video Streaming with Congestion Control

Uncompressed HD Video Streaming with Congestion Control Uncompressed HD Video Streaming with Congestion Control Ladan Gharai...University of Southern California/ISI Colin Perkins... University of Glasgow http://www.east.isi.edu/~ladan/apan.pdf Outline Goals

More information

The Future of High-Performance Networking (The 5?, 10?, 15? Year Outlook)

The Future of High-Performance Networking (The 5?, 10?, 15? Year Outlook) Workshop on New Visions for Large-Scale Networks: Research & Applications Vienna, VA, USA, March 12-14, 2001 The Future of High-Performance Networking (The 5?, 10?, 15? Year Outlook) Wu-chun Feng feng@lanl.gov

More information

Advanced Computer Networks. Flow Control

Advanced Computer Networks. Flow Control Advanced Computer Networks 263 3501 00 Flow Control Patrick Stuedi Spring Semester 2017 1 Oriana Riva, Department of Computer Science ETH Zürich Last week TCP in Datacenters Avoid incast problem - Reduce

More information

Ongoing work on NSF OCI at UNH InterOperability Laboratory. UNH IOL Participants

Ongoing work on NSF OCI at UNH InterOperability Laboratory. UNH IOL Participants Ongoing work on NSF OCI-1127228 at UNH InterOperability Laboratory Robert D. Russell InterOperability Laboratory & Computer Science Department University of New Hampshire Durham, New Hampshire

More information

Table of Contents. Cisco Introduction to EIGRP

Table of Contents. Cisco Introduction to EIGRP Table of Contents Introduction to EIGRP...1 Introduction...1 Before You Begin...1 Conventions...1 Prerequisites...1 Components Used...1 What is IGRP?...2 What is EIGRP?...2 How Does EIGRP Work?...2 EIGRP

More information

Wide-Area InfiniBand RDMA: Experimental Evaluation

Wide-Area InfiniBand RDMA: Experimental Evaluation Wide-Area InfiniBand RDMA: Experimental Evaluation Nagi Rao, Steve Poole, Paul Newman, Susan Hicks Oak Ridge National Laboratory High-Performance Interconnects Workshop August 31, 2009, New Orleans, LA

More information

A Feedback Mechanism for Network Scheduling in LambdaGrids

A Feedback Mechanism for Network Scheduling in LambdaGrids A Feedback Mechanism for Network Scheduling in LambdaGrids Pallab Datta, Sushant Sharma Computer and Computational Sciences Division Los Alamos National Laboratory Los Alamos, NM 87545 E-mail:{pallab,

More information

Internetworking Models The OSI Reference Model

Internetworking Models The OSI Reference Model Internetworking Models When networks first came into being, computers could typically communicate only with computers from the same manufacturer. In the late 1970s, the Open Systems Interconnection (OSI)

More information

Chapter 2 - Part 1. The TCP/IP Protocol: The Language of the Internet

Chapter 2 - Part 1. The TCP/IP Protocol: The Language of the Internet Chapter 2 - Part 1 The TCP/IP Protocol: The Language of the Internet Protocols A protocol is a language or set of rules that two or more computers use to communicate 2 Protocol Analogy: Phone Call Parties

More information

THE OSI MODEL. Application Presentation Session Transport Network Data-Link Physical. OSI Model. Chapter 1 Review.

THE OSI MODEL. Application Presentation Session Transport Network Data-Link Physical. OSI Model. Chapter 1 Review. THE OSI MODEL Application Presentation Session Transport Network Data-Link Physical OSI Model Chapter 1 Review By: Allan Johnson Table of Contents Go There! Go There! Go There! Go There! Go There! Go There!

More information

High-performance message striping over reliable transport protocols

High-performance message striping over reliable transport protocols J Supercomput (2006) 38:261 278 DOI 10.1007/s11227-006-8443-6 High-performance message striping over reliable transport protocols Nader Mohamed Jameela Al-Jaroodi Hong Jiang David Swanson C Science + Business

More information

Internet2 and IPv6: A Status Update

Internet2 and IPv6: A Status Update Internet2 and IPv6: A Status Update Rick Summerhill Associate Director, Network Infrastructure, Internet2 North American IPv6 Summit Santa Monica, CA 15 June 2004 Outline Internet2 Goals Participation

More information

TCP Performance. EE 122: Intro to Communication Networks. Fall 2006 (MW 4-5:30 in Donner 155) Vern Paxson TAs: Dilip Antony Joseph and Sukun Kim

TCP Performance. EE 122: Intro to Communication Networks. Fall 2006 (MW 4-5:30 in Donner 155) Vern Paxson TAs: Dilip Antony Joseph and Sukun Kim TCP Performance EE 122: Intro to Communication Networks Fall 2006 (MW 4-5:30 in Donner 155) Vern Paxson TAs: Dilip Antony Joseph and Sukun Kim http://inst.eecs.berkeley.edu/~ee122/ Materials with thanks

More information

OSI Layer OSI Name Units Implementation Description 7 Application Data PCs Network services such as file, print,

OSI Layer OSI Name Units Implementation Description 7 Application Data PCs Network services such as file, print, ANNEX B - Communications Protocol Overheads The OSI Model is a conceptual model that standardizes the functions of a telecommunication or computing system without regard of their underlying internal structure

More information

Toward a Reliable Data Transport Architecture for Optical Burst-Switched Networks

Toward a Reliable Data Transport Architecture for Optical Burst-Switched Networks Toward a Reliable Data Transport Architecture for Optical Burst-Switched Networks Dr. Vinod Vokkarane Assistant Professor, Computer and Information Science Co-Director, Advanced Computer Networks Lab University

More information

CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION

CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION 31 CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION This chapter introduces the Grid monitoring with resource metrics and network metrics. This chapter also discusses various network monitoring tools and

More information

6.1 Internet Transport Layer Architecture 6.2 UDP (User Datagram Protocol) 6.3 TCP (Transmission Control Protocol) 6. Transport Layer 6-1

6.1 Internet Transport Layer Architecture 6.2 UDP (User Datagram Protocol) 6.3 TCP (Transmission Control Protocol) 6. Transport Layer 6-1 6. Transport Layer 6.1 Internet Transport Layer Architecture 6.2 UDP (User Datagram Protocol) 6.3 TCP (Transmission Control Protocol) 6. Transport Layer 6-1 6.1 Internet Transport Layer Architecture The

More information

CS 556 Advanced Computer Networks Spring Solutions to Midterm Test March 10, YOUR NAME: Abraham MATTA

CS 556 Advanced Computer Networks Spring Solutions to Midterm Test March 10, YOUR NAME: Abraham MATTA CS 556 Advanced Computer Networks Spring 2011 Solutions to Midterm Test March 10, 2011 YOUR NAME: Abraham MATTA This test is closed books. You are only allowed to have one sheet of notes (8.5 11 ). Please

More information

EMC VPLEX Geo with Quantum StorNext

EMC VPLEX Geo with Quantum StorNext White Paper Application Enabled Collaboration Abstract The EMC VPLEX Geo storage federation solution, together with Quantum StorNext file system, enables a global clustered File System solution where remote

More information

Progress Report. Project title: Resource optimization in hybrid core networks with 100G systems

Progress Report. Project title: Resource optimization in hybrid core networks with 100G systems Progress Report DOE award number: DE-SC0002350 Name of the recipient: University of Virginia Project title: Resource optimization in hybrid core networks with 100G systems Principal investigator: Malathi

More information

Industry Perspectives on Optical Networking. Joe Berthold 28 September 2004

Industry Perspectives on Optical Networking. Joe Berthold 28 September 2004 Industry Perspectives on Optical Networking Joe Berthold 28 September 2004 CIENA LightWorks Architecture Vision Benefits for Network Operators Reduce OpEx by Process Automation Reduce CapEx by Functional

More information

BEN: Breakable Experimental Network

BEN: Breakable Experimental Network BEN: Breakable Experimental Network Nick McKeown Stanford University Presenter: Munhwan Choi Overview Requirements and motivation BEN overview Basic resources Fiber infrastructure and connectivity Functional

More information

An Experimental Analysis on OSPF-TE Convergence Time

An Experimental Analysis on OSPF-TE Convergence Time An Experimental Analysis on OSPF-TE Convergence Time S. Huang* a, K. Kitayama a, F. Cugini b, F. Paolucci c, A. Giorgetti c, L. Valcarenghi c, P. Castoldi c a Osaka University, Osaka, Japan; b CNIT, Pisa,

More information

Layer 4 TCP Performance in carrier transport networks Throughput is not always equal to bandwidth

Layer 4 TCP Performance in carrier transport networks Throughput is not always equal to bandwidth Layer 4 TCP Performance in carrier transport networks Throughput is not always equal to bandwidth Roland Stooss JDSU Deutschland GmbH Senior Consultant Data/IP Analysis Solutions Mühleweg 5, D-72800 Eningen

More information

On Network Dimensioning Approach for the Internet

On Network Dimensioning Approach for the Internet On Dimensioning Approach for the Internet Masayuki Murata ed Environment Division Cybermedia Center, (also, Graduate School of Engineering Science, ) e-mail: murata@ics.es.osaka-u.ac.jp http://www-ana.ics.es.osaka-u.ac.jp/

More information

Documents. Configuration. Important Dependent Parameters (Approximate) Version 2.3 (Wed, Dec 1, 2010, 1225 hours)

Documents. Configuration. Important Dependent Parameters (Approximate) Version 2.3 (Wed, Dec 1, 2010, 1225 hours) 1 of 7 12/2/2010 11:31 AM Version 2.3 (Wed, Dec 1, 2010, 1225 hours) Notation And Abbreviations preliminaries TCP Experiment 2 TCP Experiment 1 Remarks How To Design A TCP Experiment KB (KiloBytes = 1,000

More information

Lixia Zhang M. I. T. Laboratory for Computer Science December 1985

Lixia Zhang M. I. T. Laboratory for Computer Science December 1985 Network Working Group Request for Comments: 969 David D. Clark Mark L. Lambert Lixia Zhang M. I. T. Laboratory for Computer Science December 1985 1. STATUS OF THIS MEMO This RFC suggests a proposed protocol

More information

PLEASE READ CAREFULLY BEFORE YOU START

PLEASE READ CAREFULLY BEFORE YOU START Page 1 of 20 MIDTERM EXAMINATION #1 - B COMPUTER NETWORKS : 03-60-367-01 U N I V E R S I T Y O F W I N D S O R S C H O O L O F C O M P U T E R S C I E N C E Fall 2008-75 minutes This examination document

More information

PLEASE READ CAREFULLY BEFORE YOU START

PLEASE READ CAREFULLY BEFORE YOU START Page 1 of 20 MIDTERM EXAMINATION #1 - A COMPUTER NETWORKS : 03-60-367-01 U N I V E R S I T Y O F W I N D S O R S C H O O L O F C O M P U T E R S C I E N C E Fall 2008-75 minutes This examination document

More information

Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet

Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet Aamir Shaikh and Kenneth J. Christensen Department of Computer Science and Engineering University of South Florida Tampa,

More information

Why Your Application only Uses 10Mbps Even the Link is 1Gbps?

Why Your Application only Uses 10Mbps Even the Link is 1Gbps? Why Your Application only Uses 10Mbps Even the Link is 1Gbps? Contents Introduction Background Information Overview of the Issue Bandwidth-Delay Product Verify Solution How to Tell Round Trip Time (RTT)

More information

Next Generation Networking and The HOPI Testbed

Next Generation Networking and The HOPI Testbed Next Generation Networking and The HOPI Testbed Rick Summerhill Director, Network Research, Architecture, and Technologies, Internet2 CANS 2005 Shenzhen, China 2 November 2005 Agenda Current Abilene Network

More information

6.033 Spring 2015 Lecture #11: Transport Layer Congestion Control Hari Balakrishnan Scribed by Qian Long

6.033 Spring 2015 Lecture #11: Transport Layer Congestion Control Hari Balakrishnan Scribed by Qian Long 6.033 Spring 2015 Lecture #11: Transport Layer Congestion Control Hari Balakrishnan Scribed by Qian Long Please read Chapter 19 of the 6.02 book for background, especially on acknowledgments (ACKs), timers,

More information

Congestion Control. Tom Anderson

Congestion Control. Tom Anderson Congestion Control Tom Anderson Bandwidth Allocation How do we efficiently share network resources among billions of hosts? Congestion control Sending too fast causes packet loss inside network -> retransmissions

More information

440GX Application Note

440GX Application Note Overview of TCP/IP Acceleration Hardware January 22, 2008 Introduction Modern interconnect technology offers Gigabit/second (Gb/s) speed that has shifted the bottleneck in communication from the physical

More information

Scaling Internet TV Content Delivery ALEX GUTARIN DIRECTOR OF ENGINEERING, NETFLIX

Scaling Internet TV Content Delivery ALEX GUTARIN DIRECTOR OF ENGINEERING, NETFLIX Scaling Internet TV Content Delivery ALEX GUTARIN DIRECTOR OF ENGINEERING, NETFLIX Inventing Internet TV Available in more than 190 countries 104+ million subscribers Lots of Streaming == Lots of Traffic

More information

TCP so far Computer Networking Outline. How Was TCP Able to Evolve

TCP so far Computer Networking Outline. How Was TCP Able to Evolve TCP so far 15-441 15-441 Computer Networking 15-641 Lecture 14: TCP Performance & Future Peter Steenkiste Fall 2016 www.cs.cmu.edu/~prs/15-441-f16 Reliable byte stream protocol Connection establishments

More information

Problem 1. Multiple choice questions. Circle the answer that is the most correct.

Problem 1. Multiple choice questions. Circle the answer that is the most correct. EECS 122: Introduction to Communications Networks Midterm 1 Solutions (50 points) Wednesday March 1, 2000 Professor Sengupta Problem 1. Multiple choice questions. Circle the answer that is the most correct.

More information

USING ISCSI AND VERITAS BACKUP EXEC 9.0 FOR WINDOWS SERVERS BENEFITS AND TEST CONFIGURATION

USING ISCSI AND VERITAS BACKUP EXEC 9.0 FOR WINDOWS SERVERS BENEFITS AND TEST CONFIGURATION WHITE PAPER Maximize Storage Networks with iscsi USING ISCSI AND VERITAS BACKUP EXEC 9.0 FOR WINDOWS SERVERS BENEFITS AND TEST CONFIGURATION For use with Windows 2000 VERITAS Software Corporation 03/05/2003

More information

Computer Networks: Lab 3 Traceroute and IP Luca Bedogni

Computer Networks: Lab 3 Traceroute and IP Luca Bedogni Computer Networks: Lab 3 Traceroute and IP Luca Bedogni Department of Computer Science and Engineering University of Bologna A brief introduction We will leverage the traceroute utility It traces the route

More information

Advanced Network Design

Advanced Network Design Advanced Network Design Organization Whoami, Book, Wikipedia www.cs.uchicago.edu/~nugent/cspp54015 Grading Homework/project: 60% Midterm: 15% Final: 20% Class participation: 5% Interdisciplinary Course

More information

EEC-484/584 Computer Networks

EEC-484/584 Computer Networks EEC-484/584 Computer Networks Lecture 2 Wenbing Zhao wenbing@ieee.org (Lecture nodes are based on materials supplied by Dr. Louise Moser at UCSB and Prentice-Hall) Misc. Interested in research? Secure

More information

Reasons not to Parallelize TCP Connections for Fast Long-Distance Networks

Reasons not to Parallelize TCP Connections for Fast Long-Distance Networks Reasons not to Parallelize TCP Connections for Fast Long-Distance Networks Zongsheng Zhang Go Hasegawa Masayuki Murata Osaka University Contents Introduction Analysis of parallel TCP mechanism Numerical

More information

Introduction to Open System Interconnection Reference Model

Introduction to Open System Interconnection Reference Model Chapter 5 Introduction to OSI Reference Model 1 Chapter 5 Introduction to Open System Interconnection Reference Model Introduction The Open Systems Interconnection (OSI) model is a reference tool for understanding

More information

ASPERA HIGH-SPEED TRANSFER. Moving the world s data at maximum speed

ASPERA HIGH-SPEED TRANSFER. Moving the world s data at maximum speed ASPERA HIGH-SPEED TRANSFER Moving the world s data at maximum speed ASPERA HIGH-SPEED FILE TRANSFER Aspera FASP Data Transfer at 80 Gbps Elimina8ng tradi8onal bo

More information

CHEETAH End-Host Software Design Specification

CHEETAH End-Host Software Design Specification CHEETAH End-Host Software Design Specification 1. Introduction 1.1 Goals and Objectives CHEETAH end-host software is a part of our effort to develop the infrastructure and networking technologies to support

More information

Ultra high-speed transmission technology for wide area data movement

Ultra high-speed transmission technology for wide area data movement Ultra high-speed transmission technology for wide area data movement Michelle Munson, president & co-founder Aspera Outline Business motivation Moving ever larger file sets over commodity IP networks (public,

More information

Fundamental Questions to Answer About Computer Networking, Jan 2009 Prof. Ying-Dar Lin,

Fundamental Questions to Answer About Computer Networking, Jan 2009 Prof. Ying-Dar Lin, Fundamental Questions to Answer About Computer Networking, Jan 2009 Prof. Ying-Dar Lin, ydlin@cs.nctu.edu.tw Chapter 1: Introduction 1. How does Internet scale to billions of hosts? (Describe what structure

More information

TCP Flavors Simulation Evaluations over Noisy Environment

TCP Flavors Simulation Evaluations over Noisy Environment International Journal of Information Engineering and Applications 2018; 1(1): 11-17 http://www.aascit.org/journal/information TCP Flavors Simulation Evaluations over Noisy Environment Elsadig Gamaleldeen

More information

IST ATRIUM. A testbed of terabit IP routers running MPLS over DWDM. TF-NGN meeting

IST ATRIUM. A testbed of terabit IP routers running MPLS over DWDM. TF-NGN meeting IST 1999-20675 ATRIUM A testbed of terabit IP routers running MPLS over DWDM TF-NGN meeting 18-06-2001 http://www.alcatel.be/atrium The objectives of the presentation Present the project : objectives partners

More information

NIC TEAMING IEEE 802.3ad

NIC TEAMING IEEE 802.3ad WHITE PAPER NIC TEAMING IEEE 802.3ad NIC Teaming IEEE 802.3ad Summary This tech note describes the NIC (Network Interface Card) teaming capabilities of VMware ESX Server 2 including its benefits, performance

More information

Reliable File Transfer

Reliable File Transfer Due date Wednesday, Mar 14, 11:59pm Reliable File Transfer CS 5565 Spring 2012, Project 2 This project is worth 100 points. You may form teams of up to two students for this project. You are not required

More information

Fermilab WAN Performance Analysis Methodology. Wenji Wu, Phil DeMar, Matt Crawford ESCC/Internet2 Joint Techs July 23, 2008

Fermilab WAN Performance Analysis Methodology. Wenji Wu, Phil DeMar, Matt Crawford ESCC/Internet2 Joint Techs July 23, 2008 Fermilab WAN Performance Analysis Methodology Wenji Wu, Phil DeMar, Matt Crawford ESCC/Internet2 Joint Techs July 23, 2008 1 The Wizard s Gap 10 years and counting The Wizard Gap (Mathis 1999) is still

More information

New Approaches to Optical Packet Switching in Carrier Networks. Thomas C. McDermott Chiaro Networks Richardson, Texas

New Approaches to Optical Packet Switching in Carrier Networks. Thomas C. McDermott Chiaro Networks Richardson, Texas New Approaches to Optical Packet Switching in Carrier Networks Thomas C. McDermott Chiaro Networks Richardson, Texas Outline Introduction, Vision, Problem statement Approaches to Optical Packet Switching

More information

IEEE 1588 PTP clock synchronization over a WAN backbone

IEEE 1588 PTP clock synchronization over a WAN backbone Whitepaper IEEE 1588 PTP clock synchronization over a WAN backbone A field study comparing PTP clock synchronization accuracy against GPS external time reference in a live production WAN environment Contents

More information