Presentation is loading. Please wait.

Presentation is loading. Please wait.

Masaki Hirabaru CRL, Japan ITRC MAI BoF Kochi November 6, 2003 広帯域・高遅延ネットワークでの TCP性能計測.

Similar presentations


Presentation on theme: "Masaki Hirabaru CRL, Japan ITRC MAI BoF Kochi November 6, 2003 広帯域・高遅延ネットワークでの TCP性能計測."— Presentation transcript:

1 Masaki Hirabaru CRL, Japan ITRC MAI BoF Kochi November 6, 2003 広帯域・高遅延ネットワークでの TCP性能計測

2 Acknowledgements David Lapsley, Haystack Observatory, MIT Koyama Yasuhiro, Kashima Space Research Center, CRL Junichi Nakajima, Kashima Space Research Center, CRL Jouko Ritakari, Metsähovi Radio Observatory, HUT Katsushi Kobayashi, Information and Network Systems, CRL CRL RDNP / Tokyo XP / TransPAC / Abilene / Caltech

3 Backgrounds e-VLBI – geographically distributed observation, interconnecting radio antennas over the world Gigabit/real-time VLBI – multi-gigabit rate sampling

4 VLBI (Very Long Baseline Interferometry)

5 Motivations MIT Haystack – CRL Kashima e-VLBI Experiment on August 27, 2003 to measure UT1-UTC in 24 hours –41.54 GB CRL => MIT 107 Mbps (~50 mins) 41.54 GB MIT => CRL 44.6 Mbps (~120 mins) –RTT ~220 ms, UDP throughput 300-400 Mbps However TCP ~6-8 Mbps (per session, tuned) –BBFTP with 5 x 10 TCP sessions to gain performance HUT – CRL Kashima Gigabit VLBI Experiment -RTT ~325 ms, UDP throughput ~70 Mbps However TCP ~2 Mbps (as is), ~10 Mbps (tuned) -Netants (5 TCP sessions with ftp stream restart extension) They need high-speed / real-time / reliable / long-haul high-performance, huge data transfer.

6 Purpose Ensure >= 1 Gbps end-to-end performance in high bandwidth-delay product networks –to support for networked science applications –to help operations in finding a bottleneck –to evaluate advanced transport protocols (e.g. Tsunami, SABUL, HSTCP, FAST, XCP, ikob)

7 Contents Advanced TCP evaluation on TransPAC/Internet2 How to isolate host and implementation specific constraints TCP monitoring with web100 Advanced TCP evaluation in a laboratory Summary and future work

8 Kwangju Busan 2.5G Fukuoka Korea 2.5G SONET KOREN Taegu Daejon 10G 0.6G1Gx2 QGPOP Seoul XP Genkai XP Kitakyushu Tokyo XP Kashima 0.1G Fukuoka Japan 250km 1,000km 2.5G TransPAC 9,000km 4,000km Los Angeles Cicago New York MIT Haystack HUT 10G 1G APII/JGN Abilene 0.1G Helsinki 2.4G Stockholm 0.6G 2.4G GEANT Nordunet funet Koganei 1G 7,000km Indianapolis I2 Venue 1G 10G 100km server (general) server (e-VLBI) Abilene Observatory: servers at each NOC CMM: common measurement machines Network Diagram for TransPAC/I2 Measurement (Oct. 2003) 1G x2 sender receiver Mark5 Linux 2.4.7 (RH 7.1) P3 1.3GHz Memory 256MB GbE SK-9843 PE1650 Linux 2.4.22 (RH 9) Xeon 1.4GHz Memory 1GB GbE Intel Pro/1000 XT Iperf UDP ~900Mbps (no loss)

9 TransPAC/I2 #1:Reno (Win 64MB)

10 TransPAC/I2 #1:Reno (better case)

11 TransPAC/I2 #1:Reno (worse case)

12 TransPAC/I2 #1: Reno (10 mins)

13 TransPAC/I2 #2:High Speed (Win 64MB)

14 TransPAC/I2 #2:High Speed (better case)

15 TransPAC/I2 #2:High Speed (worse case)

16 TransPAC/I2 #2: High Speed (60 mins)

17 TransPAC/I2 #1: Reno (Win 12MB)

18 TransPAC/I2 #2: High Speed (Win 12MB)

19 *Testing FAST TCP over Abilene by Shalunov, Internet2 Meeting, Oct. 2003 Path4: Pittsburgh to Atlanta, RTT=26.9ms

20 FASTLinux throughput loss queue STCPHSTCP 30min Room for mice ! HSTCP * TCP comparison (dummynet 800M, 120ms) from FAST TCP presented IETF Vienna July 2003

21 1 st Step: Tuning a Host with UDP Remove any bottlenecks on a host –CPU, Memory, Bus, OS (driver), … Dell PowerEdge 1650 (*not enough power) –Intel Xeon 1.4GHz x1(2), Memory 1GB –Intel Pro/1000 XT onboard PCI-X (133Mhz) Dell PowerEdge 2650 –Intel Xeon 2.8GHz x1(2), Memory 1GB –Intel Pro/1000 XT PCI-X (133Mhz) Iperf UDP throughput 957 Mbps –GbE wire rate: headers: UDP(20B)+IP(20B)+EthernetII(38B) –Linux 2.4.22 (RedHat 9) with web100 –PE1650: TxIntDelay=0

22 Evaluating Advanced TCPs Reno (Linux TCP, web100 version) –Ack: w=w+1/w, Loss: w=w-1/2*w HighSpeed TCP (included in web100) –Ack: w=w+a(w)/w, Loss: w=w-1/b(w)*w FAST TCP (binary, provided from Caltech) –w=1/2*(w_old*baseRTT/avgRTT+α+w_current) Limited Slow-Start (included in web 100) Note: Differences in sender side only

23 Web100 (http://www.web100.org) A kernel patch for monitoring/modifying TCP metrics in Linux kernel Alpha 2.3 for Linux 2.4.22, released September 12, 2003

24 2 nd Step: Tuning a Host with TCP Maximum socket buffer size (TCP window size) –net.core.wmem_max net.core.rmem_max (64MB) –net.ipv4.tcp_wmem net.tcp4.tcp_rmem (64MB) Driver descriptor length –e1000: TxDescriptors=1024 RxDescriptors=256 (default) Interface queue length –txqueuelen=100 (default) –net.core.netdev_max_backlog=300 (default) Interface queue descriptor –fifo (default) MTU –mtu=1500 (IP MTU) Iperf TCP throughput 941 Mbps –GbE wire rate: headers: TCP(32B)+IP(20B)+EthernetII(38B) –Linux 2.4.22 (RedHat 9) with web100 Web100 (incl. High Speed TCP) –net.ipv4.web100_no_metric_save=1 (do not store TCP metrics in the route cache) –net.ipv4.WAD_IFQ=1 (do not send a congestion signal on buffer full) –net.ipv4.web100_rbufmode=0 net.ipv4.web100_sbufmode=0 (disable auto tuning) –Net.ipv4.WAD_FloydAIMD=1 (HighSpeed TCP) –net.ipv4.web100_default_wscale=7 (default) FAST TCP from Caltech –Linux 2.4.20 –net.ipv4.tcp_vegas_cong_avoid=1 –net.ipv4.tcp_vegas_pascing=1 net.ipv4.tcp_vegas_fast_converge=1 –net.ipv4.tcp_vegas_alpha=100 net.ipv4.tcp_vegas_beta=120 net.ipv4.tcp_vegas_gamma=50

25 Test in a laboratory (1) – no bottleneck Packet Sphere ReceiverSender L2SW (12GCF) Bandwidth 1Gbps Buffer 0KB Delay 88 ms Loss 0 GbE/SX GbE/T PE 2650PE 1650 #0: Reno => Reno –net.ipv4.WAD_IFQ=0 (congestion signal on buffer full) #1: Reno => Reno #2: High Speed TCP => Reno #3: FAST TCP => Reno 2*BDP = 11MB #0, 1, 2: Data obtained on sender #3: Data obtained on receiver

26 Laboratory #0: Reno (no bottleneck)

27 Laboratory #0: Reno (60 mins)

28 Laboratory #1: Reno (no bottleneck)

29 Laboratory #1: Reno (60 mins)

30 Laboratory #2: High Speed (no bottleneck)

31 Laboratory #2: High Speed (60 mins)

32 Laboratory #3: FAST (no bottleneck)

33 Laboratory #3: FAST (60 mins)

34 Test in a laboratory (2) – with bottleneck Packet Sphere ReceiverSender L2SW (12GCF) Bandwidth 800Mbps Buffer 256KB Delay 88 ms Loss 0 GbE/SX GbE/T PE 2650PE 1650 #4: Reno => Reno #5: High Speed TCP => Reno #6: FAST TCP => Reno 2*BDP = 11MB #4, 5: Data obtained on sender #6: Data obtained on receiver

35 Laboratory #4: Reno (800Mbps)

36 Laboratory #4: Reno (startup)

37 Laboratory #4: High Speed (800Mbps)

38 Laboratory #5: High Speed (startup)

39 Laboratory #5: High Speed (180-360)

40 Laboratory #5: High Speed (Win 12MB)

41 Laboratory #5: High Speed (Limited slow-start)

42 Laboratory #7: FAST (800Mbps)

43 Laboratory #7: FAST (startup)

44 Test in a laboratory (3) – with bottleneck Packet Sphere ReceiverSender L2SW (12GCF) Bandwidth 100Mbps Buffer 256KB Delay 88 ms Loss 0 GbE/SX GbE/T PE 2650PE 1650 #7: Reno => Reno #8: HighSpeed TCP => Reno #9: FAST TCP => Reno #10: Reno => Reno (100M shaping) 2*BDP = 11MB #7, 8, 10: Data obtained on sender #9: Data obtained on receiver

45 Laboratory #7: Reno (100Mbps)

46 Laboratory #7: Reno (10 mins)

47 Laboratory #8: High Speed (100Mbps)

48 Laboratory #8: High Speed (10 mins)

49 Laboratory #9: FAST (100Mbps)

50 Laboratory #9: FAST (10 mins)

51 Laboratory #10: Reno (100Mbps shaping)

52 Summary and Future work No significant differences under no bottlenecks –Packet loss With bottleneck, FAST would work better in case RTT increase can be detected –RED on the bottleneck –Queue length on the bottleneck –Signal from the bottleneck or from the receiver Appropriate window size to avoid loss –Autotuning? Mobile

53 Questions? See http://www2.crl.go.jp/ka/radioastro/index.html for VLBI


Download ppt "Masaki Hirabaru CRL, Japan ITRC MAI BoF Kochi November 6, 2003 広帯域・高遅延ネットワークでの TCP性能計測."

Similar presentations


Ads by Google