Sat
Sun
Mon
Tue
Saturday, October 21, 2000
Topic/Presenter
Recordings
Full Abstract

This talk will discuss the mechanisms used in optical networks to provide highly available services, and their applicability to IP networks. Topics to be covered include:

  1. Protection
    • What is it?
    • Protection vs. Dynamic Rerouting
  2. Protection topologies
    • Linear
    • Ring
    • Mesh
    • Local vs. Global Repair
  3. Protection Hierarchy
    • Optical Layer
    • SONET
    • MPLS/IP
  4. Use in SONET/SDH Networks
    • APS Switching
    • UPSR

    Speakers
    Jonathan Sadler, Tellabs
Recordings
Full Abstract

Topics covered in this tutorial include:

  • IS-IS basics
  • Addressing in IS-IS
  • IS-IS database
  • SPF (Dikjstra) operation
  • Recent IS-IS enhancements
  • MPLS TE using IS-IS

The session also includes information about configuring IS-IS routing on Cisco routers, and configuring MPLS traffic engineering using IS-IS on Cisco routers. IS-IS Configurations

NEW: Other router vendors are welcome to participate in this session by supplying IS-IS configs comparable to those shown by Abe.

Redback Configs Example 

 An Example of Redback ISIS configuration October 12, 2000 [email protected] (1) Redback router IS-IS setup Router "sunnyvale-c1" has 3 IS-IS interfaces and are connected to IS-IS instances "Optical-Backbone" on fe1/1, to instance "IP-Metro-SanJose" on sr2/1 and fe3/1. IS-IS Authentication is applied to all IS-IS packets on instance "Optical-Backbone" with hmac-md5, and authentication is applied only on IIH packets on instance "IP-Metro-SanJose". IS-IS level-2 prefixes are leaked into level-1 with the prefix-list policy on "Optical-Backbone". The IS-IS "Optical-Backbone" prefixes are redistributed into instance "IP-Metro-SanJose" on level-1 with a route-map policy applied. IS-IS instance "IP-Metro-SanJose" is running with "wide" metric style. ----------------------------------------------------------------------- (2) show IS-IS portion of the configuration [local]sunnyvale-c1#show configuration isis Current configuration: context local ! interface fe1/1 ip address 10.1.1.1/24 ip router isis Optical-Backbone isis circuit type level-2-only ! interface sr2/1 ip address 10.3.1.1/24 ip router isis IP-Metro-SanJose isis circuit type level-1 isis authentication key-chain bar ! interface fe3/1 ip address 10.7.1.1/16 ip router isis IP-Metro-SanJose isis authentication key-chain foo type simple ! router isis Optical-Backbone net 47.0001.1111.1111.1111.00 interarea-distribute l2-to-l1 prefix-list filter-100/8 authentication key-chain foo ! router isis IP-Metro-SanJose net 49.0002.1111.1111.1111.00 redistribute isis Optical-Backbone level-1 route-map metro-sanjose metric-style wide [local]sunnyvale-c1# ----------------------------------------------------------------------- (3) show IS-IS interface information [local]sunnyvale-c1#show isis interface [local]sunnyvale-c1#show isis interfaces IS-IS interface(s) for tag Optical-Backbone: Interface L P State Level-1-DR Level-2-DR Metric fe1/1 2 Up sunnyvale-c1.01 10 Total IS-IS Interface(s): 1 IS-IS interface(s) for tag IP-Metro-SanJose: Interface L P State Level-1-DR Level-2-DR Metric sr2/1 1 Up sunnyvale-c1.01 10 fe3/1 3 Up sunnyvale-c1.02 sunnyvale-c1.02 10 Total IS-IS Interface(s): 2 [local]sunnyvale-c1# ----------------------------------------------------------------------- (4) show IS-IS interface detail information [local]sunnyvale-c1#sh isis interfaces detail IS-IS interface(s) for tag Optical-Backbone: fe1/1 Up, Ckt level: 2, lan, IP address: 10.1.1.1/24, Ckt id: 0x01 Level Adjs Priority Hello Hold Auth Blocked Metric 2 1 64 5 24 10 Total IS-IS Interface(s): 1 IS-IS interface(s) for tag IP-Metro-SanJose: sr2/1 Up, Ckt level: 1, lan, IP address: 10.3.1.1/24, Ckt id: 0x02 Level Adjs Priority Hello Hold Auth Blocked Metric 1 1 64 9 24 md5 10 fe3/1 Up, Ckt level: 3, lan, IP address: 10.7.1.1/16, Ckt id: 0x03 Level Adjs Priority Hello Hold Auth Blocked Metric 1 1 64 4 24 simple 10 2 1 64 5 24 simple 10 Total IS-IS Interface(s): 2 [local]sunnyvale-c1# ----------------------------------------------------------------------- (5) show IS-IS adjacency information [local]sunnyvale-c1#show isis adjacency IS-IS Adjacenc(ies) for tag Optical-Backbone: SystemId Interface Lvl State Holdtime SNPA Uptime santa-cruz-b1 fe1/1 2 Up 22 00d0.b714.1512 01:18:33 Total IS-IS Adjacenc(ies): 1 IS-IS Adjacenc(ies) for tag IP-Metro-SanJose: SystemId Interface Lvl State Holdtime SNPA Uptime palo-alto-c2 sr2/1 1 Up 18 0090.27af.4269 00:06:15 4444.4444.4444 fe3/1 2 Up 22 0030.949f.cb00 00:06:16 4444.4444.4444 fe3/1 1 Up 22 0030.949f.cb00 00:06:19 Total IS-IS Adjacenc(ies): 3 [local]sunnyvale-c1# ----------------------------------------------------------------------- (6) show IS-IS redistributed or leaked prefixes [local]sunnyvale-c1#sh isis routes redistribute IS-IS Redistributed route(s) for tag Optical-Backbone, on Level-1 Prefix L Type Source Metric M-Type Summarized 100.0.0.0/24 1 Leak isis 15 Int 100.0.1.0/24 1 Leak isis 15 Int 100.0.2.0/24 1 Leak isis 15 Int 100.0.3.0/24 1 Leak isis 15 Int 100.0.4.0/24 1 Leak isis 15 Int Total IS-IS Redistributed Route(s) in level-1 for tag Optical-Backbone: 5 IS-IS Redistributed route(s) for tag IP-Metro-SanJose, on Level-1 Prefix L Type Source Metric M-Type Summarized 100.0.0.0/24 1 Ext isis 20 Ext Total IS-IS Redistributed Routes in level-1 for tag IP-Metro-SanJose: 1 IS-IS Redistributed route(s) for tag IP-Metro-SanJose, on Level-2 Prefix L Type Source Metric M-Type Summarized 10.3.1.0/24 2 Leak isis-intf 10 Int 10.7.0.0/16 2 Leak isis-intf 10 Int Total IS-IS Redistributed Routes in level-2 for tag IP-Metro-SanJose: 2 [local]sunnyvale-c1# ----------------------------------------------------------------------- (7) show IS-IS route summary information [local]sunnyvale-c1#show isis routes summary IS-IS route(s) summary for tag Optical-Backbone: Route Type Level-1 Level-2 Summarize(L1/L2) L2-to-L1 Leak IS-IS Route 0 11 - 0 Redistribute 0 0 0/0 Inter-area 5 0 0/0 Summary Address 0 0 0/0 IS-IS interface routes: 1 Route leaking prefix lists: filter-100/8(l2-to-l1) IS-IS route(s) summary for tag IP-Metro-SanJose: Route Type Level-1 Level-2 Summarize(L1/L2) L2-to-L1 Leak IS-IS Route 3 0 - 0 Redistribute 1 0 0/0 Inter-area 0 2 0/0 Summary Address 0 0 0/0 IS-IS interface routes: 2 Redistributed protocols: isis [local]sunnyvale-c1# ----------------------------------------------------------------------- (8) show one IS-IS prefix This prefix 100.0.3.0/24 is learn through level-2 with nexthop of 10.1.1.2 on interface fe1/1. This route is from LSP 2222.2222.2222.00-00 of seq# 0x24 from router santa-cruz-b1. That lsp was last arrived 4:35 ago from interface fe1/1. [local]sunnyvale-c1#show isis route 100.0.3.0/24 IS-IS prefix for tag Optical-Backbone: Prefix Level Metric Interface Nexthop 100.0.3.0/24 2 15 fe1/1 10.1.1.2 Is sourced from LSP(s): LSP ID Seq # System Name Arrive(ago) Interface(from) 2222.2222.2222.00-00 0x24 santa-cruz-b1 00:04:35 fe1/1 [local]sunnyvale-c1# ----------------------------------------------------------------------- (9) some debug IS-IS information This output captured the debug output on isis adjacency events for both send and receive packets on interface fe1/1 with packet source systemID of 2222.2222.2222: [local]sunnyvale-c1#debug isis adjacency filter interface fe1/1 system-id ? String System ID (XXXX.XXXX.XXXX) [local]sunnyvale-c1#$adjacency filter interface fe1/1 system-id 2222.2222.2222 [local]sunnyvale-c1# Oct 11 04:19:34: %ISIS-7-ADJ: rcvd L2 LAN IIH from 00d0.b714.1512 on intf fe1/1 Oct 11 04:19:36: %ISIS-7-ADJ: send L2 LAN IIH on intf fe1/1 Oct 11 04:19:42: %ISIS-7-ADJ: rcvd L2 LAN IIH from 00d0.b714.1512 on intf fe1/1 Oct 11 04:19:43: %ISIS-7-ADJ: send L2 LAN IIH on intf fe1/1 [local]sunnyvale-c1#no debug isis all [local]sunnyvale-c1#

Speakers
Abe Martey, Cisco Systems
Sunday, October 22, 2000
Topic/Presenter
Full Abstract

ISPs are regulated by a complex set of rules governing their creation, retention, and disclosure of customer communications and transactions. This session provides an overview of the specific rules in the controlling Federal statute, the Electronic Communications Privacy Act, and the legal consequences of failing to comply.

Speakers
Mark Eckenwiler, Department of Justice
Mark Eckenwiler is Senior Counsel in the Computer Crime and Intellectual Property Section, Criminal Division, U.S. Department of Justice. His areas of responsibility include federal wiretap law, computer search and seizure, and online investigations.

An Internet veteran for 15 years, Mark has written and spoken widely on such issues as anonymity and free speech, e-mail stalking laws, Internet jurisdiction, electronic privacy, and the Fifth Amendment implications of cryptographic keys. His articles have appeared in The National Law Journal, Legal Times, American Lawyer, Civil RICO Report, Internet World, and NetGuide. Mark holds an A.B. cum laude from Harvard in History and Literature, an M.A. in Classics (Ancient Greek) from Boston University, and a J.D. cum laude from New York University School of Law.
Recordings
Full Abstract

In this talk, we expand on our earlier NANOG presentation with an exploration of what roles inter-domain topology and routing policy play in the process of delayed Internet routing convergence. In our previous talk, we demonstrated that the Internet lacks effective inter-domain path fail-over with backbone routers requiring tens of minutes to reach a consistent view of the network topology after a fault. We also presented a theoretic, upper factorial bound on BGP convergence computation.

Based on analysis from the experimental injection and measurement of several hundred thousand additional inter-domain routing faults, this talk provides a tighter, more realistic/expected bound on BGP convergence. We describe how specific routing policies and Internet topologies impact the time for convergence. Finally, we describe and provide probable explanations for significant measured variations in the convergence behaviors of different Internet service providers.

Speakers
Abha Ahuja, Internap/Merit
Craig Labovitz, Microsoft Research/Merit
Srinivasan Venkatachary, Microsoft Research
Roger Wattenhofer, Microsoft Research
Full Abstract

The purpose of this panel is to discuss desired route filtering capabilities of services providers (both panel members and attendees), as well as to express perceived shortcomings with the different components required to effectively support large numbers of prefixes (e.g., routing registries, vendor support, management and provisioning, etc.). It's relatively well understood that service providers need to employ customer ingress prefix filters. However, the feasibility and usefulness of inter-provider filters has yet to be realized. Once fully supported, these filters may potentially be used for other applications as well.

Speakers
Moderator - Danny McPherson, Amber Networks.
Panelist - Alex Bligh, GXN
Panelist - Randy Bush, Verio
Panelist - Vijay Gill, MFN
Panelist - Tony Tauber, Genuity
Monday, October 23, 2000
Topic/Presenter
Recordings
Full Abstract

This talk provides an overview of the continued evolution of the MAEs, including a technical description of an additional service type that is being rolled out on MAE-ATM. The status of MAE-FDDI and current work on new exchange fabrics will also be covered.

Speakers
Thomas Bechly, WorldCom
Tom has been with WorldCom, MCI, and BT as software developer, network enginger, architect, and engineering manager for various Internet and data services.
Recordings
Full Abstract

Public Internet exchange points have traditionally been built around LAN switches or PVC switches (ATM and Frame Relay.) New communication technologies are becoming readily available, such as MPLS, IP-over-optical, DWDM, and others.

In this session, panelists and the audience will discuss:

  • How can this new technology be used to improve public exchange points and build new ones?

  • What services to ISPs really need from exchange points?

  • What architectures best serve these needs?

Panelists represent exchange point providers, ISPs, and researchers.

Speakers
Moderator - Steve Feldman, Sigma Networks
Panelist - Mike Hughes, LINX
Panelist - Lane Patterson, Equinix
Panelist - Bill St. Arnaud, CANARIE
Panelist - Paul Vixie, PAIX/MFN
Full Abstract

Amber Networks
Full Abstract

Internet Service Providers interconnect in peering and transit relationships in order to provide their customers with access to the global Internet. Previous studies of Internet Operations (see Peering Decision Tree, NANOG 19) in this area have highlighted key challenges facing ISPs that seek additional peering-based (non-transit) interconnections. One challenge is that the peering process is undocumented and the negotiations are veiled under non-disclosure agreements. As a result, peering discussions are hampered with confusion and misunderstandings. This led to the creation of the Peering Simulation Game, first introduced at NANOG 19, which has proven to be effective in highlighting key peering issues, and stimulating discussion amongst the audience and the players of the simulation.

This presentation and the resulting discussion are based upon operations research surrounding the current practices documented in the white paper Peering in the US, based upon the three phases of Internet peering:

  • Identification of potential peer,
  • Initial discussions, and
  • Implementation
The report is based upon interviews with about 50 ISP Peering Coordinators, and has been presented at US and non-US forums as well (initially presented at NANOG Peering BOF in Montreal, IEPG in Oslo, APRICOT in Seoul, Interconnect Accounting and Billing in London, and the ITU meeting in Rio de Janeiro) and has been updated to incorporate feedback from the European and Asia Pacific arenas.

Speakers
Bill Norton, Equinix
As Co-Founder and Director of Business Development at Equinix, Bill Norton focuses his attention on building strategic relationships among companies participating at the Internet Business Exchanges. Previously, he was the Chair of NANOG and Manager of the Internet Engineering Group at Merit, leading a variety of national and international network research and operations projects.
Full Abstract

A discussion this summer on the NANOG list prompted Sassaman to consider ways that the ISP community can benefit and benefit from a robust PGP keyserver network. This BOF provides a forum for discussing the integration of PGP services into the Internet as a function that NSPs could provide.

In order for a robust keyserver network to exist, the cooperation of multiple providers will be necessary. If done correctly, we can all benefit from this. The existence of the program PGP is one of the things that makes the Internet what it is: a haven for free thought. If we can give the public anything that will make it easier to use this tool, we will have done something good.

Currently there are numerous problems with the keyserver network; the most obvious is frequent downtime in keyservers. The most serious is the lack of synchronization (and thus lack of propagation of certificate revocations) between keyservers. Both of these problems are due to the fact that most keyservers are run by private individuals, educational institutions, and other entities who do not have the level of resources available to ensure the existence of a stable, reliable, and robust network of keyservers.

We will discuss these problems, and then talk about the ways that a cooperative effort between network service providers could rectify this situation. Sassaman will also give a brief introduction to the PGP web of trust model, for those who aren't familiar with it.

In this BOF we hope to take the first steps to change things for the better!

Speakers
Len Sassaman, PGP Security
Len Sassaman is one of the leading experts on the PGP encryption software, and has worked with it since 1993. He left his career in the ISP field to take a position as software engineer at PGP Security, working on PGP key server development. Additionally he is the technical lead for the FreeCert project, and founder of Known Safe, Inc. His primary focus is PKI architecture.
Full Abstract

Speakers
Susan R. Harris, Merit Network
Cleve Mickles, AOL Time Warner
Full Abstract

We are actively involved in ongoing efforts in the IETF, Optical Internetworking Forum, and Optical Domain Service Interconnect to standardize IP-centric control architecture and protocols for optical networks. Our objective is to discuss the following issues and get service providers' feedback on them:

  • IP-centric control optical architectures for optical networks and their impact on layer 3 networks,

  • Functional requirements of optical layer control functions, namely, link management, topology discovery, routing, and signaling, and the ways they are different from similar functions in IP networks,

  • Standards activities in these areas (e.g. MPLS, OSPF, UNI, etc.), implementation, and deployment issues.

Speakers
Bala Rajagopalan, Tellium
Bala Rajagopalan is a lead architect at Tellium, Inc. He is responsible for Tellium's IP over optical and optical internetworking architectures. Prior to joining Tellium, Dr. Rajagopalan was with AT&T Bell Laboratories, Bellcore, and NEC C&C Research Laboratories, where he developed protocols and algorithms for wireless data networking, Internet routing and multicast, QoS, traffic engineering, and IP switching. Rajagopalan has published numerous papers and delivered many invited lectures on topics covering his areas of interest, and he has filed patents relating to wireless services and internetworking, and optical networks. He is also very active in standardization activities and he has made significant contributions to the IETF, the ATM Forum and the Optical Interworking Forum. Rajagopalan received his B. Tech degree from the Indian Institute of Technology, and his M.S and Ph.D degrees from the University of Illinois at Urbana-Champaign, all in Computer Science.

Debanjan Saha, Tellium
Debanjan Saha is the principal architect of Tellium's IP centric service management platform. Previous to his tenure at Tellium, he spent several years at IBM Research and Lucent Bell Labs, where he designed and developed protocols for IP routers, multi-service switches, and UNIX servers. He is actively involved with various standards bodies, most notably IETF and OIF. He also serves as editor of international journals and magazines, and technical committee member of workshops and conferences. Dr. Saha is a notable author of numerous technical articles on various networking topics, and is a frequent speaker at academic and industrial events. He holds a B.Tech. degree from IIT, India, and M.S. and Ph.D. degrees from the University of Maryland at College Park, all in Computer Science.
Full Abstract

There is currently a worldwide shortage of 10Gbs optical components for manufacturing SONET OC-192 transponders. As a result, OC-192 interfaces for routers and DWDM equipment are very expensive and difficult to obtain. Carriers have estimated that about 75% of all SONET interfaces connect to other equipment in the same facility, with a link length of under 300m. The Optical Internetworking Forum (OIF) has been working for over a year on Very Short Reach (VSR) OC-192 interfaces which trade link length for lower cost and better manufacturability.

Several technical approaches have been considered, including parallel optics, VCSEL lasers, relaxed optical parameters, and coarse WDM. The presentation will present in detail the characteristics of the leading approaches in order to show their advantages and disadvantages. It will probably make sense to adopt more than one. Input is sought from NANOG attendees on the best criteria to use in deciding which to adopt. The criteria of interest include maximum reach, type of fiber required, difference in cost, and difference in availability date.

This work is being done by the OIF's Physical and Link Layer Working Group (PLL WG). The OIF develops and publishes Implementation Agreements, rather than official standards. The first VSR Implementation Agreement is expected to be approved this fall. VSR line cards have already been demonstrated by multiple router vendors.

Speakers
Russ Tuck, Pluris
Russ Tuck chairs the OIF's Physical and Link Layer Working Group, which is developing VSR Implementation Agreements. He is Systems Architect at Pluris Terabit Network Systems, where he helps lead the development of scalable and reliable core routers. He previously architected scalable database servers at Pyramid Technology, and scalable supercomputers at MasPar. He earned a Ph.D. in Computer Science from Duke University and carried out his dissertation research at UNC Chapel Hill.
Full Abstract

  • Network Reliability
    P.J. Aduskevicz, ATT/FCC Network Reliability & Interoperability Council (NRIC)

  • Wireless Developments
    Dale Hatfield, FCC Office of Engineering and Technology

  • Broadband Deployment
    Rebecca Dorch, FCC Office of Engineering and Technology

  • Cable Open Access
    Robert Cannon, FCC Office of Plans & Policy

For more than 30 years, the FCC has pursued a policy of unregulation with regards to data processing and computer networks. Nevertheless, the decisions of the Federal Communications Commission can have dramatic effects on the future of the Internet. This panel looks at some of the most noteworthy issues before the FCC that affect Internet development.

The Network Reliability and Interoperability Council, an industry group which has advised the FCC on telecommunications reliability, has, for the first time, expanded its vision to include a one year voluntary trial monitoring the reliability of non-traditional networks such as the Internet. Mobile devices such as wireless palms and browsers on cell phones have created pressure for making new wireless spectrum available for technological advances. Increased Internet usage has increased demand for bandwidth and created a new broadband market to the residence; the FCC's recent Section 706 Report examined whether advanced broadband services are being deployed to all Americans in a reasonable and timely manner. Finally, an intense controversy has been brewing concerning whether cable services should have an obligation to open facilities to some form of non-discriminatory access by unaffiliated ISPs; the FCC has recently initiated an inquiry to look at the issue of "open access."

Speakers
Robert Cannon, FCC
Robert Cannon is Senior Counsel for Internet Issues in the Office of Plans and Policy of the Federal Communications Commission. Prior to this position, he was Deputy Director of the FCC's Y2K Task Force. He is also the Founder of the Washington Internet Project, a pro-bono project dedicated to promoting awareness of and participation in federal regulatory developments that affect the Internet.

Mr. Cannon moderates the Cybertelecom-l listserv and edits the e-newsletter CybertelecomNews. His article on the Communications Decency Act was published in the Federal Communications Law Journal, cited before the Supreme Court in Reno v. ACLU and republished in an anthology on the First Amendment. He is a Legal Columnist for Boardwatch Magazine and has spoken at ISPCON, INET, APRICOT, ISPF, Computers, Freedom and Privacy, and Internet World. He was the creator and co-chair of the Federal Communication Bar Association's Online Communications Committee. In 1993, he completed a judicial clerkship with Judge Steffen Graae in D.C. Superior Court.
Recordings
Full Abstract

Speakers
Marcus Thomas, FBI
Recordings
Full Abstract

SURFnet, the national network for research and higher education in The Netherlands, is engaged in the GigaPort Project with the aim of building the next generation Internet in The Netherlands. SURFnet is handling all traffic for its research and higher education users, as opposed to the model in North America, where universities have to have a commercial ISP to handle commodity traffic and can have an Abilene and/or vBNS connection for just research traffic.

One of the sub-projects in the GigaPort Project is the building of SURFnet5, which is undertaken in partnership with Telfort, a wholly-owned subsidiary of British Telecom, and Cisco Systems. The final network of SURFnet5 will consist of a backbone at two locations in Amsterdam and 15 concentrator locations connected to the backbone using IP-over-DWDM with POS framing at 20 Gbit/s (2x OC-192c). The backbone and 14 concentrator locations will be ready for pre-production during the end of the calendar year 2000. Features such as MPLS fast reroute (FRR) will be key in this network. Key topics in this talk are:

  • Inside the GigaPort Project we see applications and application development as an integral part of the project. Much work is being done on video over IP, with live multicasting of the three Dutch public channels over our infrastructure, with MPEG-1 quality.

  • MPLS TE fast reroute will be elaborated on since SURFnet5 will be 10G IP-over-lambda without any restoration facilities on the DWDM layer. We have to rely on FRR to successfully build and operate our network. Besides MPLS TE fast reroute, we are heavily involved in MPLS VPN testing. I plan on giving the latest results of the testing.

  • Lessons learned during the Pilot phase of SURFnet5 in which we connected four Cisco 12016 (GSR+) routers located at four different locations in The Netherlands using dark fiber and DWDM sections at OC-48c. IP performance testing at multiple gigabit/s level using two clusters of Suns and Dells and FRR testing are amongst the topics touched upon. Also the plans for further extending this network are presented, e.g., the plan to move to 80 Gbit/s in the core during calendar year 2002.

Speakers
Erik-Jan Bos, SURFnet
Erik-Jan Bos is head of the network services department of SURFnet, and is responsible for building SURFnet5 in the context of the GigaPort Project. Erik-Jan has been involved in building IP networks since early 1989.
Full Abstract

Speakers
Susan R. Harris,, Merit Network
Tuesday, October 24, 2000
Topic/Presenter
Recordings
Full Abstract

The FBI's National Infrastructure Protection Center (NIPC) has observed a recent increase in DDoS attacks, with many newly discovered amplifier systems in use. This talk will describe recent tool deployments and ask for attendee input about:

  • How the FBI can quickly notify ISP's when the NIPC receives information about amplifiers/masters, etc., which are on either the ISP's own networks or those of their customers.

  • How the FBI can help get information about DDoS activity to ISP's, which can then be further disseminated to their customers

The NIPC's goal is not only to limit damage to victims, but also to help ISP's protect their own networks and those of their customers.

Presentation slides.

Speakers
Steve Schmidt, FBI
Steve Schmidt, formerly an engineer with American Information Systems, is Chief of the Special Technologies and Applications Unit at the NIPC.
Recordings
Full Abstract

The primary purpose of this presentation is to list and analyze the currently available Global Server Load Balancing (GSLB) methods. A new GSLB technique along with the testing results is presented that is shown to be the most efficient solution in some network setups. Various GSLB network scenarios are considered, ranging from a set of few GSLB'ed servers to global content delivery ASP network setups. What GSLB method should be used in what network setup is also discussed.

Speakers
Alex Kit, Winstar Communications
Alex Kit is a Senior Internet Systems Architect at Winstar Communications. He is currently involved in engineering scalable, highly-redundant, fault-tolerant, geographically distributed network and systems architectures to support Internet-based applications. These include, but are not limited to, Web hosting, ASP, Internet core services (mail, DNS, news, IRC, etc.).

Dmitri Krioukov, Nortel
Dima Krioukov is a lead technologist in the Internet Data Center/ASP Network Infrastructure Practice Group of Nortel Networks. His current projects include research and analysis of Global Load Balancing and [BGP/MPLS] VPN solutions as well as addressing their applicability issues in the context of content delivery networks. Dima has been involved in IDC network infrastructure engineering since 1998.
Recordings
Full Abstract

This presentation describes an approach to IP multicast performance measurement on the NSF's very-high-speed Backbone Network Service (vBNS). Using OC-12c attached workstations acting as multicast senders or receivers distributed throughout the vBNS backbone, we create arbitrary topologies, generate synthetic IP multicast traffic and measure loss encountered. The presentation includes packet loss results as a function of time and router hop count and an analysis of the join latency incurred.

The multicast tests described in the presentation have been used by the vBNS for the past year as a diagnostic tool for monitoring the state of the production multicast netork. In addition, the presentation will contain timely information for network operators interested in the potential capabilities of multicast, scaling properties and testing methods.

Speakers
Robert Beverly, MCI WorldCom
Robert Beverly is a senior engineer with WorldCom's Advanced Internet Technology group in Ashburn, Virginia, where he has worked for the past two years on the very high performance Backbone Network Service (vBNS). His primary interests include network performance, statistics and visualization. He received his B.S. in computer engineering from the Georgia Institute of Technology in 1996.
Recordings
Full Abstract

This talk describes a tool for characterizing the TCP behavior of a remote host on the Internet. The specific goal in building this tool was to answer the question, "What fraction of web servers use NewReno instead of Reno or Tahoe TCP congestion control mechanisms, for TCP connections with non-Sack-enabled clients? The more general goal was to provide a tool for efficiently probing the TCP congestion control behaviors of remote hosts in the Internet. We report on both the tool and our experimental results.

For more information about the new tool, see:

http://www.aciri.org/tbit/">http://www.aciri.org/tbit/

For presentation slides, please see:

http://www.icir.org/padhye/nanog-tbit.pdf">http://www.icir.org/padhye/nanog-tbit.pdf

Speakers
Sally Floyd, Aciri
Jitendra Padhye, Aciri
Recordings
Full Abstract

I am very interested in finding out anything about what fraction of routers in the Internet do or don't use active queue management (e.g., RED), and what experiences with RED have been like. While it is not so hard to run tests to find out about the TCP congestion control behaviors of web servers, I don't know of any way to find out, from the outside, about the current queue management behavior of routers. And it would be helpful to know more than I do now ... Please contact me at the meeting or send email if you'd like to exchange information about this topic.

http://www.icir.org/floyd/talks/NANOG_Oct00.pdf">Presentation slides

Speakers
Sally Floyd, Aciri
Sally Floyd is a research scientist at ACIRI, the AT&T Center for Internet Research at ICSI. She has worked on a range of congestion-control-related issues, including TCP congestion control mechanisms, active queue management, and Explicit Congestion Notification (ECN).
Recordings
Full Abstract

Speakers
Cengiz Alaettinoglu, Packet Design
Cengiz Alaettinoglu is a member of the Technical Staff at Packet Design Inc. He was previously at the USC Information Sciences Institute, where he worked on the Routing Arbiter project. Cengiz co-defined the Routing Policy Specification Language along with the protocols to enable a distributed, secure routing policy system.

Van Jacobson, Packet Design
Van Jacobson is Chief Scientist at Packet Design Inc. He was previously Chief Scientist at Cisco Systems. Prior to his tenure at Cisco, he spent 25 years at Lawrence Berkeley Laboratory as a Group Leader for the Network Research Group.

Haobo Yu, Packet Design
Haobo Yu is a member of the Technical Staff at Packet Design. His research interests include web cache consistency protocols and application-level routing. He received his Ph.D. in 2000 from USC.
Recordings
Full Abstract

Speakers
John Scudder, Cisco Systems
Download the full event calendar