1. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL1 Report to ICFA August 10, 1999 Matthias Kasemann, FNAL

2. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 2 Related URL’s lICFA-SCIC Homepage èhttp://www.hep.net/ICFA/index.html Cern -> Scientific Committees -> ICFA -> ICFA Standing Committee on International Connectivity lICFA-NTF Homepage èhttp:/nicewww.cern.ch/~davidw/icfa/icfa-ntf.html lICFA-NTF July’98 Report èhttp://nicewww.cern.ch/~davidw/icfa/July98Report.html

3. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 3 ICFA meeting, Vancouver, 1998 lICFA received the final report of the Networking Task Force (ICFA-NTF). lDecision: create a Standing Committee on Interregional Connectivity (ICFA-SCIC). lCommittee members represent major HEP user communities and laboratories. lFocus should be on intercontinental connectivity (see charge).

4. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 4 Charge to ICFA-SCIC: lMake recommendations to ICFA concerning the connectivity between America, Aisia and Europe. èAs part of the process of developing these recommendations, the committee should smonitor traffic, skeep track of technology developments, speriodically review forecasts of future bandwidth needs, and sprovide early warning of potential problems. lCreate subcommittees when necessary to meet the charge. lThe chair of the committee should report to ICFA once a year, at its joint meeting with laboratory directors.

5. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 5 ICFA-SCIC membership: lThe chair is appointed directly by ICFA.  lEach of the major user laboratories, CERN, DESY, FERMILAB, KEK and SLAC, should appoint one member each.  lECFA, DPF jointly with IPP, and ACFA, should appoint two members each.  lICFA will appoint one member from the Russian federation and one member from South America. 

6. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 6 ICFA-SCIC membership: lThe representatives from the laboratories are: Manuel Delfino, (CERN), Michael Ernst (DESY), Kasemann (Fermi) (chair), Yukio Karita (KEK), Richard Mount (SLAC). lThe North American user representatives are: Harvey Newman (USA), Dean Karlen (Canada). lFor Russia: Alexei Morozow (ITEP) l ECFA has nominated: Frederico Ruggieri (INFN Frascati), Denis Linglin (IN2P3, Lyon). l ACFA has nominated: Prof. Rongsheng Xu (Computer Center, IHEP China) Prof. HwanBae Park (Korea University) l For South America: Sergio F. Novaes (University de S.Paulo)

7. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 7 ICFA-SCIC meetings: lApril 15. - 16. at FNAL. èMain topics: sreview charge to SCIC, sreview work of ICFA-NTF ( a lot of it overlaps with SCIC charge), sdefine priorities and projects, sorganize work (and subgroups). lVideo Conference on July 6, 1999. èTopics: supdate on status of network connectivity sworking group plans saction items

8. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 8 Recommendations of the ICFA NTF lRecommendations concerning Inter-continental links: èICFA should encourage the provision of some considerable extra bandwidth, especially across the Atlantic èICFA participants should make concrete proposals, (such as recommendation to increase bandwidth across the Atlantic, approach to QoS, co-operation with other disciplines and agencies, etc.) èThe bandwidth to Japan needs to be upgraded èIntegrated end-to-end connectivity is primary requirement, to be emphasized to continental ISPs, and academic and research networks

9. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 9 Topics: lThere was a discussion on the reality of QoS today: èhow to administer it èhow to take advantage of it for HEP applications sbulk data transfer service to be done at low utilization periods, smove away from applications that demand low latency such as telnet èthere is a need for network aware applications for: sinteractive network connections sdistinguish bulk data transfer from low latency traffic sIP-telephony, Voice over IP scollaborative tools

10. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 10 Topics (2): lStatus & future directions of US research networks  The network environment for HEP research in the US continues to improve in virtually all areas  US HEP facilities work well on ESnet  University access improving rapidly with emerging I2 networks  Acceptable Use Policy (AUP) issues complicate things

11. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 11 Canada - Germany(Desy): poor There are severe trans- atlantic performance problems which need to be addressed. They need inter- national cooperation. ICFA-SCIC (Michael Ernst, DESY) will discuss with DFN. unusable poor ~ ok excellent

12. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 12 Canada - Germany(Desy): poor

13. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 13 S. America / Brazil lThe needs are to increase bandwidth. lNeed A&R dedicated links, even better if had just HEP links. lFor a D0 farm they need 20kbps/PC or for 30 PC need 600kbps (total available on US link is 2000kbps). lThe main HEP partners are CERN & FNAL. lUnclear how much fiber/infrastructure is being put in place to the US. èthere are at least 2 international consortia Oxygen, Global Crossing are going into Brazil. èFunding is the main problem and limits international connectivity. lThe telecomm industry is being privatized, expect more competition.

14. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 14 Germany lEurope connectivity improved with TEN-155, lExpect an upgrade of A&R net in Germany next year. lProblem areas: èN. America (2*OC3 won't help much); èRussia now & Japan in future. lDFN now does ICMP traffic shaping especially at International exchanges. lHoping to be able to use Differentiating Services to provide managed bandwidth for improved performance to HEP sites in N. America.

15. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 15 Recent History of the US-CERN Link l CERN/US/France/Canada/UN-WHO Consortium lOctober 1996 - August 1997 è Upgraded leased digital CERN-US line: 2.048 Mbps è Set-up of monitoring tools and traffic control è Start Deployment of VRVS a Web-based videoconferencing system lSeptember 1997 - April 1999 è Upgraded leased CERN-US line to 2 X 2.048 Mbps; Addition of a backup and “overflow” leased line at 2.048 Mbps (total 6 Mbps) to avoid saturation in Fall 1998 è Production deployment of VRVS software in the US and Europe (to 1000 hosts by 4/99; Now 1300). è Set-up of CERN-US consortium rack at Perryman to peer with ESnet and other international nets è Test of QoS features using new Cisco software and hardware

16. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 16 Current Dev. for the US-CERN Link l October 1998 - September 1999 èMarket survey and selection of Cable&Wireless as ISP. èBegan Collaboration in Internet2 applications and network developments. èMove to C&W Chicago PoP, to connect to STARTAP. èFrom April 1999, set-up of a 12 Mbps ATM VP/VBRnrt circuit between CERN and C&W PoP è9/99: Transatlantic upgrade to 20 Mbps September 1st, coincident with CERN/IN2P3 link upgrade è7/99: Begin organized file transfer service to “mirror” Babar DST data from SLAC to CCIN2P3/Lyon

17. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 17 Bandwidth Growth Observation/Prediction lTechnology Tracking and Cost Model will be performed by the ICFA SCIC Committee

18. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 18 Japan site report - Yukio Karita lIssues: èvBNS is reluctant to peer with Japan HENP at STAR-TAP èNACSIS - Europe line is saturated: swill be upgraded from 2 Mbps to 30 Mbps on October 1, 1999. s3-4 Mbps for KEK-CERN ATM PVC will be provided then. sBoth of this done with Japanese $$’s. sEuropean funding can increase the bandwidth.

19. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 19 Remote Collaborations: VRVS videoconferencing lVRVS CERN-Caltech development (1995- l21 reflectors Running in U.S. Europe and Asia èSwitzerland: CERN (2) èItaly: CNAF Bologna èUK: Rutherford Lab èFrance: IN2P3 Lyon, Marseilles èGermany: Heidelberg Univ. èFinland: FUNET èSpain: IFCA-Univ. Cantabria èRussia: Moscow State Univ., Tver. U. èU.S: s Caltech, LBNL, SLAC, FNAL, s ANL, BNL, Jefferson Lab. èDoE HQ Germantown Academia Sinica. Taiwan - Asia : Academia Sinica. Taiwan South America: CeCalcula.Venezuela - South America: CeCalcula.Venezuela

20. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 20 VRVS: Widespread and Strong Support from the Research and Education Communities l HENP Community. Hosts registered from: CMS, Atlas, Alice, Lhc-b, Aleph, NA48, NA49, NA50, AMS, Aleph, Babar, RHIC, CDF, Ceres/NA45, Chorus, Delphi, DESY/ZDV, H1, CEBAF, KLOE, KTeV, L3, Minos, Soudan2, OPAL, PHENIX, STAR, SpEcTrE, WA95, WA98, ZEUS, etc … l Strong interest from others Research Communities èInternet2/UCAID (University Corporation for Advance Internet Development) sTed Hans, Director, Application Development Internet2: l“..The Internet2 Community sees VRVS as the model for providing a highly functional video-enabled collaboration service for research and education…” sI2-DV (Internet2 Digital Video) Initiative recognized that VRVS is a uniquely suitable foundation for development and deployment of its applications. (http://i2dv.nwu.icair.org) sP. Galvez is a member of the I2-DV (Internet2 Digital Video) steering committee

21. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 21 Example: 9 Participants, CERN(2), Caltech, FNAL(2), Bologna (IT), Roma (IT), Milan (IT), Rutherford(UK)

22. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL22 SCIC monitoring WG Active Internet Monitoring Activities Les Cottrell – SLAC For the SCIC-WG

23. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 23 Overview of Mechanism 19 Monitor sites, 10 countries 1300 monitor-remote-site pairs 379 unique hosts, 27 countries Measure response, jitter, loss, reachability Data goes back > 4 years 1 Million probes of Internet/day Treats Internet as black box Uses existing infrastructure (ping) Low cost, well understood Ping WWW HEPNRC SLAC Monitor Remote

24. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 24 Deployment in HEP lOver 50% of HEP collaborator sites are explicitly monitored as remote sites by PingER èAtlas (37%), BaBar(68%), Belle(23%), CDF(73%), CMS (31%), D0(60%), Zeus (35%), Aleph, Delphi, Opal, L3 (43%) èCreated focussed PingER pages for BaBar, CDF, D0 lRemainder represented by beacon sites èSelected to represent countries/R&E nets èAbout 50 beacon sites in 27 countries

25. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 25 Performance Trends Bandwidth TCP < 1460/(RTT * sqrt(loss))

26. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 26 Problem areas Germany bad with Canada &.edu yet good with ESnet Russia (W) bad to everywhere yet good with Esnet China poor with most UK varies with time, bandwidth increased by factor of ~80 in 4 years Peering is critical, avoid congestion points, reduce number of NSPs, reliance on your NSP

27. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 27 Overview & Future directions of Internet Monitoring in HEP lAction Items:  Need to add India and S. America to the Beacon sites.   Calibrate the formula  TCP BW < (MSS/RTT)* (1/sqrt(loss)) with measurements within the HEP community, e.g. CERN-SLAC, SLAC-LBNL etc.  HEPNRC will redouble its efforts to get more heavily involved with PingER once again.  Increase efforts to gather and archive traceroutes between major sites

28. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 28 Network monitoring conclusions lPerformance is getting better lESnet is well provisioned lWithin R&E networks things are good lTransoceanic, needs special care lPeering is critical lMonitoring bare network èIf one wants to go beyond ping to start to understand network aware applications and understand the barriers to higher throughput it will need to involve the applications people. This would be less global than PingER, i.e. more focussed on point-to-point particular links and applications. This is not part of the monitoring group's responsibilities. èReport on Survey & tracking of HEP network aware applications, QoS, prototyping work in the area of: sdistributed databases, data transfer, collaborative tools

29. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 29 SCIC Summary: Working Groups lRequirements Analysis Working Group (Harvey Newman & Matthias Kasemann) èProposed action scompare expected requirements for 2000 to reality scomment on "network aware" applications, R&D prototyping effort sReport to ICFA in Spring 2000 lTechnology, status, cost & development/expectations for HEP (Richard Mount, Michael Ernst, David Williams) èKeep track of technology trends: sreview networking activities and initiatives, sreport on pilot projects, shave active participation of ICFA-SCIC members in projects

30. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 30 SCIC Summary lNeed small focussed task force to address particularly bad performance areas:  Japan-US, Japan - Europe: some relief on the horizon  Canada-Germany (getting Germany to STARTAP) lDocument Status of Remote regions  get reports from India, Pakistan  UK - N. America: invite UK person to report on connectivity.

31. August 10, 1999 ICFA_SCIC Report #1 Matthias Kasemann, FNAL 31 SCIC Summary lReports to ICFA planed: èSummer 1999: monitoring results, short report on activity, report on short term bottlenecks èSpring 2000 / 2002 /... every 2 years: monitoring results, recommendations for long term situation èSpring 2001 / 2003 /... every 2 years: monitoring results, update report on requirements èFirst report to ICFA on Feb 10,11 2000, (report ready by October, 1999) è3rd Meeting: November 13, 1999 CERN at which we derive recommendations for the ICFA report.