KEK Network Qi Fazhi 2004.8. KEK SW L2/L3 Switch for outside connections Central L2/L3 Switch A Netscreen Firewall Super Sinet Router 10GbE 2 x GbE IDS.

Slides:



Advertisements
Similar presentations
Martin Suchara, Ryan Witt, Bartek Wydrowski California Institute of Technology Pasadena, U.S.A. TCP MaxNet Implementation and Experiments on the WAN in.
Advertisements

Congestion Control and Fairness Models Nick Feamster CS 4251 Computer Networking II Spring 2008.
TCP transfers over high latency/bandwidth network & Grid TCP Sylvain Ravot
LOGO Transmission Control Protocol 12 (TCP) Data Flow.
CCNA3: Switching Basics and Intermediate Routing v3.0 CISCO NETWORKING ACADEMY PROGRAM Switching Concepts Introduction to Ethernet/802.3 LANs Introduction.
Fundamentals of Computer Networks ECE 478/578
TCP Congestion Control Dina Katabi & Sam Madden nms.csail.mit.edu/~dina 6.033, Spring 2014.
Iperf Tutorial Jon Dugan Summer JointTechs 2010, Columbus, OH.
Ahmed El-Hassany CISC856: CISC 856 TCP/IP and Upper Layer Protocols Slides adopted from: Injong Rhee, Lisong Xu.
Advanced Computer Networking Congestion Control for High Bandwidth-Delay Product Environments (XCP Algorithm) 1.
Congestion Control Tanenbaum 5.3, /12/2015Congestion Control (A Loss Based Technique: TCP)2 What? Why? Congestion occurs when –there is no reservation.
High-performance bulk data transfers with TCP Matei Ripeanu University of Chicago.
1/20 Introduction Outline Statistical Multiplexing Inter-Process Communication Network Architecture Performance Metrics Implementation Issues.
DataTAG Meeting CERN 7-8 May 03 R. Hughes-Jones Manchester 1 High Throughput: Progress and Current Results Lots of people helped: MB-NG team at UCL MB-NG.
Internet Bandwidth Measurement Techniques Muhammad Ali Dec 17 th 2005.
CS335 Networking & Network Administration Tuesday, April 20, 2010.
TCP. Learning objectives Reliable Transport in TCP TCP flow and Congestion Control.
Ch. 28 Q and A IS 333 Spring Q1 Q: What is network latency? 1.Changes in delay and duration of the changes 2.time required to transfer data across.
All rights reserved © 2006, Alcatel Accelerating TCP Traffic on Broadband Access Networks  Ing-Jyh Tsang 
Layer 2 Switch  Layer 2 Switching is hardware based.  Uses the host's Media Access Control (MAC) address.  Uses Application Specific Integrated Circuits.
Lecture 1, 1Spring 2003, COM1337/3501Computer Communication Networks Rajmohan Rajaraman COM1337/3501 Textbook: Computer Networks: A Systems Approach, L.
Network Performance Measurement Atlas Tier 2 Meeting at BNL December Joe Metzger
1 ESnet Network Measurements ESCC Feb Joe Metzger
Chapter 4. After completion of this chapter, you should be able to: Explain “what is the Internet? And how we connect to the Internet using an ISP. Explain.
Experiences in Design and Implementation of a High Performance Transport Protocol Yunhong Gu, Xinwei Hong, and Robert L. Grossman National Center for Data.
Brierley 1 Module 4 Module 4 Introduction to LAN Switching.
These materials are licensed under the Creative Commons Attribution-Noncommercial 3.0 Unported license (
POSTECH DP&NM Lab. Internet Traffic Monitoring and Analysis: Methods and Applications (1) 4. Active Monitoring Techniques.
ECEN “Internet Protocols and Modeling” Course Materials: Papers, Reference Texts: Bertsekas/Gallager, Stuber, Stallings, etc Grading (Tentative):
UDT: UDP based Data Transfer Yunhong Gu & Robert Grossman Laboratory for Advanced Computing University of Illinois at Chicago.
workshop eugene, oregon What is network management? System & Service monitoring  Reachability, availability Resource measurement/monitoring.
1 Lecture 14 High-speed TCP connections Wraparound Keeping the pipeline full Estimating RTT Fairness of TCP congestion control Internet resource allocation.
Transport over Wireless Networks Myungchul Kim
High TCP performance over wide area networks Arlington, VA May 8, 2002 Sylvain Ravot CalTech HENP Working Group.
High-speed TCP  FAST TCP: motivation, architecture, algorithms, performance (by Cheng Jin, David X. Wei and Steven H. Low)  Modifying TCP's Congestion.
HighSpeed TCP for High Bandwidth-Delay Product Networks Raj Kettimuthu.
TCP Trunking: Design, Implementation and Performance H.T. Kung and S. Y. Wang.
Vertical Optimization Of Data Transmission For Mobile Wireless Terminals MICHAEL METHFESSEL, KAI F. DOMBROWSKI, PETER LANGENDORFER, HORST FRANKENFELDT,
1 Network Measurement Summary ESCC, Feb Joe Metzger ESnet Engineering Group Lawrence Berkeley National Laboratory.
NET100 Development of network-aware operating systems Tom Dunigan
Masaki Hirabaru NICT Koganei 3rd e-VLBI Workshop October 6, 2004 Makuhari, Japan Performance Measurement on Large Bandwidth-Delay Product.
1 Capacity Dimensioning Based on Traffic Measurement in the Internet Kazumine Osaka University Shingo Ata (Osaka City Univ.)
T. S. Eugene Ngeugeneng at cs.rice.edu Rice University1 COMP/ELEC 429/556 Introduction to Computer Networks Principles of Congestion Control Some slides.
TCP transfers over high latency/bandwidth networks Internet2 Member Meeting HENP working group session April 9-11, 2003, Arlington T. Kelly, University.
Jennifer Rexford Fall 2014 (TTh 3:00-4:20 in CS 105) COS 561: Advanced Computer Networks TCP.
Performance Engineering E2EpiPEs and FastTCP Internet2 member meeting - Indianapolis World Telecom Geneva October 15, 2003
Interconnect Networks Basics. Generic parallel/distributed system architecture On-chip interconnects (manycore processor) Off-chip interconnects (clusters.
Chapter 11.4 END-TO-END ISSUES. Optical Internet Optical technology Protocol translates availability of gigabit bandwidth in user-perceived QoS.
NET100 Development of network-aware operating systems Tom Dunigan
TCP continued. Discussion – TCP Throughput TCP will most likely generate the saw tooth type of traffic. – A rough estimate is that the congestion window.
TCP transfers over high latency/bandwidth networks & Grid DT Measurements session PFLDnet February 3- 4, 2003 CERN, Geneva, Switzerland Sylvain Ravot
Company LOGO Network Management Architecture By Dr. Shadi Masadeh 1.
TCP Traffic Characteristics—Deep buffer Switch
© Janice Regan, CMPT 128, CMPT 371 Data Communications and Networking Congestion Control 0.
Final EU Review - 24/03/2004 DataTAG is a project funded by the European Commission under contract IST Richard Hughes-Jones The University of.
Day 13 Intro to MANs and WANs. MANs Cover a larger distance than LANs –Typically multiple buildings, office park Usually in the shape of a ring –Typically.
Connect communicate collaborate Performance Metrics & Basic Tools Robert Stoy, DFN EGI TF, Madrid September 2013.
iperf a gnu tool for IP networks
Realization of a stable network flow with high performance communication in high bandwidth-delay product network Y. Kodama, T. Kudoh, O. Tatebe, S. Sekiguchi.
Networking between China and Europe
Transport Protocols over Circuits/VCs
Wide Area Networking at SLAC, Feb ‘03
Network Performance Measurement
TCP Congestion Control
Beyond FTP & hard drives: Accelerating LAN file transfers
TCP flow and congestion control
Anant Mudambi, U. Virginia
Review of Internet Protocols Transport Layer
Presentation transcript:

KEK Network Qi Fazhi

KEK SW L2/L3 Switch for outside connections Central L2/L3 Switch A Netscreen Firewall Super Sinet Router 10GbE 2 x GbE IDS GbE KEK GW 2 x GbE KEK SW L2/L3 Switch for outside connections Central L2/L3 Switch B Netscreen Firewall 2 x GbE 4 x GbE NOOD Switch to other places NOOD Switches to Buildings 4 x GbE 10GbE Tskuba Center 10GbE Other Univ. HEP Lab. GbE 4 x GbE up to: 14 Vlan, Based MAC Static & DHCP(for wireless) IDS GbE

KEK ’ s firewall architecture L2/L3 Switch for outside connections Central L2/L3 Switch Firewall 10GbE 10GbE/GbE 10GbE 2 x GbE Security monitor High Throughput Access Route Regular flows IDS GbE

KEK universities/institutes in Asia-Pacific universities in Japan SLAC, FNAL, BNL, CERN, DESY, SDSC,..... SuperSINET APAN/Transpac APAN

Since SuperSINET’s links to US/EU SuperSINET NII’s router at NewYork OC48 x 4 Gèant Abilene ESnet OC48 GbE ISP ( 10Gbps ) MAN LAN 10GbE Switch POS Channel 10GbE GbE x 3 IEEAF Atlantic SLAC is on ESnet CERN is on GEANT

WAN Performance Measurement and Monitor -- Study and Work at KEK Qi Fazhi

Study WAN Performance Terminology WAN Performance Measurement Tools How to get larger throughput

Terminology / Elements be relationship to WAN Performance Bandwidth –Physical bandwidth, or capacity (C) –Available bandwidth (A) Throughput –Maximum throughput –Achievable throughput Latency / RTT

TERMINOLOGY Latency Packet Loss Bandwidth: the speed that a network element can forward traffic. (is independent of end hosts and protocol type) –Physical/capacity:maximum number of bits per second a network element can transfer(is determined by the slowest network element along the path) –Available:the capacity minus utilization over a given time interval. Throughput: amount of data that is successfully sent from one host to another via a network (may be limited by every component along the path from source host to destination host) –Maximum throughput –Achievable throughput

Network Performance Measurement Tools Iperf/Netperf: tools for measuring end-to-end TCP/UDP performance pipechar: hop-by-hop bottleneck analysis tool pchar: hop-by-hop performance measurement tool traceroute: lists all routers from current host to remote host

How to get larger throughput Decrease Latency / RTT Packet Loss Choose the best pathRouting coordination Larger Throughput TCP Tuning ? Latency / RTT Packet Loss Parallel stream TCP

TCP Tuning Why? –Bandwidth increases –TCP shortage How? –Modify the TCP Windows Size –Modify the System Buffer Size –…… –Some New Protocol: FAST

TCP TCP is adaptive It is constantly trying to go faster It slows down when it detects a loss How much it sends is controlled by windows When it sends is controlled by received ACK’s(or timeouts)

Summary of techniques to maximize TCP WAN throughput congestion window (CWND):TCP uses it to determine how many packets can be sent at one time, The larger the congestion window size, the higher the throughput. TCP “ slow start ” and “ congestion avoidance ” algorithms determine the size of the congestion window The maximum congestion window is related to the amount of buffer space that the kernel allocates for each socket (System) For each socket, there is a default value for the buffer size, which can be changed by the program using a system library call just before opening the socket. Kernel enforced maximum buffer size. Why Tuning?

Bandwidth*Delay Product Bandwidth * Delay = number of bytes in flight to fill path The optimal buffer size is twice the bandwidth*delay product of the link( recover from errors ) buffer size = 2 * bandwidth * delay = bandwidth * RTT bandwidth :get it from pipechar or pchar pipecharpchar RTT:get it from Ping Example:ping time is 50 ms ; the end-to-end network consists of all 100 BT Ethernet and OC3 (155 Mbps).then TCP buffers should be.05 sec * (100 Mbits / 8 bits) = 625 KBytes most of today UNIX OS's by default have a maximum TCP buffer size of only 256 KB! So we should do some tuning …… buffer Windows throughput BW*Delay

Important Points About TCP Throughput congestion window : determine how many packets can be sent at one time,the larger the congestion window size, the higher the throughput. the size of the congestion window : slow start ( constantly trying to go faster ) and congestion avoidance ( slows down when it detects a loss ) maximum congestion window : is related to the amount of buffer space that the kernel allocates for each socket buffer size : 2*Bandwidth*Delay Products ( BDP ) = bandwidth * RTT Summary

Routing coordination KEK IHEP Before Change Status –Go through Russia and USA KEK->IHEP: KEK->Sinet->Abilene->RBNet->CSTnet->IHEP IHEP->KEK: IHEP->CSTnet-> RBNet ->Abilene ->Sinet ->KEK –RTT is very large After Change Status –KEK->IHEP: KEK->SINet->APAN-JP->CERnet->NSFC->CSTnet->IHEP –IHEP->KEK (not changed): IHEP->CSTnet-> RBNet ->Abilene ->Sinet ->KEK –For the reason of not changing IHEP->KEK route,the RTT is still large,but there are some changes in the throughput

KEK-IHEP Route Tuning KEK2505 Sinet2907 Abilene APAN_JP 7660 RBnet5568 Dargon_tap 9407 Cernet 4538 NSFC 9406 CSTNet 7497 AS1239 AS9405 AS3356 Level 3 Sprint APAN_JP&APAN_CN Link IHEP3460

Parallel TCP Stream KEK->IHEP One TCP StreamParallel TCP Streams

WAN Performance Measurement/Monitor Target –Packet loss & latency –Throughput Tools Mechanism WORK

Test PC Data Remote WWW Traceroute/Ping iperf netperf HTTP Ping/iperf/netperf Mechanism (Base on www) CMDResult

System Structure WAN Performance Measurement and Monitor WAN Performance MeasurementWAN Performance Monitor Performance data collect Data File Performance data analysis Performance measurement resultPerformance measurement command

The System Interface

Measurement: (RTT & Packet Loss/traceroute & ping)

Traceroute and Ping Result

Traceroute and Ping Result with graph ms

Measurement: Throughput with Iperf

Measurement Result: Throughput with Iperf

Measurement: Throughput with Netperf

Measurement Result: Throughput with Netperf

Monitor: RTT and Packet loss with ping

Monitor Result: RTT and Packet loss with ping

Monitor: Throughput with iperf

Monitor Result: Throughput with iperf

README INSTALL –creat a user account: monitor –cd ~monitor –tar xvf monitor.tar –edit /usr/local/apache/conf/httpd.conf –edit the ~monitor/ping_list.txt and ~monitor/iperf_list.txt –edit the /etc/crontab –

Useful links