- Test - lists.nanog.org

RE: NANOG/ARIN in Chicago
by Pete Templin 10 Apr '03

10 Apr '03

I'm sure I'd feel terrorized if my plane had been stranded there. Pete Templin IP Network Engineer TexLink Communications (210) 892-4183 pete.templin(a)texlink.com -----Original Message----- From: Mark Radabaugh [mailto:mark@amplex.net] Sent: Thursday, April 10, 2003 3:03 PM To: nanog(a)merit.edu Subject: Re: NANOG/ARIN in Chicago > A little quick on the gun with the "Terrorist" label? Is everyone you > disagree with in public office a "Terrorist"? > Nope - I absolutely mean terrorist. When you show up at 1:00am with Bulldozers and destroy public property without telling anyone what you are about to do what would you call it? What do you think you would get called if you did it? Mark Radabaugh Amplex (419) 720-3635

1 0

RE: The Paradox of Commoditization
by St. Clair, James 10 Apr '03

10 Apr '03

Jack Bates wrote: *I have yet to see someone swear by their own life that the new *technologies will meet the uptimes of the legacy. The fact is, in the *telco world, they don't. Good point, but I think what Gordon *may be* saying is that part of the reason the above is true is the degree monopolies are struggling to keep their legacy systems. I would argue those folks who have paid (handsomely) for *pure* transitions to new technologies have seen benefits; it is the persistent hybrids of new and legacy that complicate service. *Even businesses that require time sensitive, guaranteed communications *don't trust the new technology whole heartedly. Hmm, I would disagree. More and more critical processes (such as RTU controls at utilities) are going to new technologies, with the ROI being less expense of special technicians and remote controls. Dell "bet the farm" on new technologies for e-commerce and helped turn PCs into a commodity. Just a thought... Jim

3 2

Dual node vs "Reliable IP" Architecture
by Pete Kruckenberg 10 Apr '03

10 Apr '03

http://www.nspllc.com/New%20Pages/Reliable%20IP%20Nodes.pdf Argues that by going from a dual-router POP design to a single redundant router configuration, I can reduce annual downtime costs by 93% (?) and reduce CAPEX and OPEX (seems logical). Also mentions that the single redundant router can get closer to the ever-elusive 5-9's. Anyone doing this in their network? Is there validity in the claims in this white paper? Anyone looked at the Alcatel product that apparently funded this paper? Pete.

4 4

Would someone...
by McBurnett, Jim 10 Apr '03

10 Apr '03

Hi all, Would someone with a clue at Charter Communications drop me a line off list? I have an issue I need resolved and even your NOC is not helping me.... Thanks, Jim

1 0

Re: The Paradox of Commoditization
by Gordon Cook 10 Apr '03

10 Apr '03

Hi David Bigge and Peter Galbavy, Peter K and James St Clair David's point first..... of course people will pay for convenience and will pay for phone service. phone service is not going away and bankruptcy won't make the lecs go away. bankruptcy will transform them (at least i hope it will). My point was not intended to be JUST VoIP....although VoIP is, i believe, the current most powerful force driving commoditization and the rate at which it is being adopted will make a bad situation much worse for the LECs. The future, in addition to a commodity end-to-end IP net where everything rides IP as an application is in areas like linux and open source. What fascinates me is that I see interesting evidence that IBM gets the hypothesis I laid out. Everyone knows the huge amounts IBM has invested in linux and i was also aware that IBM was positioning itself in the user services and education area. At VON i got first hand evidence of this. There was an IBMer from the IBM Institute for Business Value in Chicago on one of the panels. Afterwards I grabbed him and we talked one on one for 90 minutes. ABSOLUTELY AMAZING....... if there is such a thing as an Isenberg-Googin-Cook school of thought, he is a full fledged member... IBM ....big huge.... former mainframe.... now decentralized. Plunging head long into customer education..... IBM GETS IT! Customer education, consulting, and services is the area that will stay solid and likely grow. Virtually everything else will shrink. Will it shrink by another million in the US? It could but it will likely take a few more years. My point in the essay was to encourage folk to think about the commoditization dynamic. I had not thought about it in quite the same way myself before. no one can predict what the time table will be. Telecom and IT won't go away, but they will never in the future operate with anything like the same economics and employment patterns that they did as recently as 2 or 3 years ago. Peter K's point As for sounding like David Isenberg. Absolutely. And Roxane Googin as well. the net paradox URL cited is as much googin as isenberg. Apparently at the spring VON 2002 Isenberg, Googin, Anders Comstadt of Stokab the municipal dark fiber net in stockholm which is making a profit, and Tim Horan of CIBC had a panel in a breakout room with 15 people in the audience. This year at 5pm april one the same four were in the main auditorium talking to an audience of about 800. the wisdom is spreading. Peter G's point, sorry -- it was anything but a very thinly veiled jingoistic plug for 'Made in USA' It is my best attempt to report what i see. And that is that " made in the USA" is to some extent history...like it or not. This part is reality and inexorable. It was a lament for the stupidity of my own government in the form of the FCC and the current congress critters for not understanding how the duopoly approach to broadband is cutting our own throats from the point of view leaving us with a second rate and expensive infrastructure upon which our future economic activity depends. The same commoditization is affecting European telcos for sure. The question is whether the EC will deal with the dynamics more intelligently than we have in the US. Question: Is telecom a sick industry? If it is, then you regulate it like the FCC has been doing in the US. If on the other had you see telecom as a sick underlying infrastructure on which economic growth depends then you better think of doing treating the industry in a way different than you would if you were just trying to preserve share holder value. Jim St Clair's point largely agree. Think big picture cost point of view. To deliver voice as a telco what does your current infrastructure COST, and what is its complexity? Give cost, complexity and debt service how can you compete against another player who is delivering its service via IP over gig E and doesn't need to buy all that complicated legacy switching gear. And increasingly large enterprises are going to themselves deliver telecom services. Case in point is IBM which on march 3rd i believe announced a huge project to globally convert to voice over IP. To do so they will rip out token ring in all their offices and install ethernet at a minimum estimated cost of $100 million. the calculation is that they will save more than that on their phone bills and via use of SIP proxies they can get new features of tailorability in voice service that cannot be gotten from the PSTN many thanks - now back to your regularly scheduled operational issues -- ============================================================= The COOK Report on Internet, 431 Greenway Ave, Ewing, NJ 08618 USA 609 882-2572 (PSTN) 703 738-6031 (Vonage) cook(a)cookreport.com Subscription info & prices at http://cookreport.com/subscriptions.shtml The Paradox of Commoditization - Studies in VoIP and Telecom Economics, April -June 2003, 128 pages available at http://cookreport.com/12.01-02.shtml and http://cookreport.com/12.03.shtml =============================================================

2 1

RE: Dual node vs "Reliable IP" Architecture
by Jim Deleskie 10 Apr '03

10 Apr '03

While I try not let fork lifts drive around much in CO's I've been responsible for :), I have seen more then once, a device ( I'll not mention the vendors, yes plural) smoke out before, not real fires, but enough to destroy a box. Now does the frequency of such events justify the cost of have dual routers everywhere, ( 2 times in 1000's of devices over almost 10 years)I'm not sure, I'll leave that up to upper management. I just build according to the dollars I have available. -Jim On Thu, 10 Apr 2003, Neil J. McRae wrote: > > Anyone doing this in their network? Is there validity in the > > claims in this white paper? Anyone looked at the Alcatel > > product that apparently funded this paper? > > I'd believe the numbers in this, what I don't believe yet > though is that there are products available that true give > the same level of redundancy that having two boxes does... I've always found that these types of papers and products always miss one big area of failure, at least in my experience. What happens when the highly redundant device is skewered by a fork lift? Yes, I've had this happen. At least if you have a dual router config, and separate those routers physically, you have a chance of surviving such problems. -- Brandon Ross AIM: BrandonNR VP Operations ICQ: 2269442 Sockeye Networks

1 0

The Paradox of Commoditization
by Gordon Cook 10 Apr '03

10 Apr '03

The Paradox Of Commoditization Trying To Save What Is Inevitably Lost, We Lose What Could Otherwise Be Gained Introduction The monopoly control of customers by Legacy networks is destroying the economic benefits that could be obtained from the on going pervasive and inexorable commoditization of telecom and information technology. We face a paradox. While we have eyes, we cannot see. We act as though we could wish away what is happening to new products and prices. But the fact is that the on-going commoditization of technology cannot be undone. Products will continue to get better but they will also continue to fall in price. In the face of these dynamics jobs will melt away. The only growth in the industry will be come from a variety of education, customer support, strategic evaluation and consulting positions. The only additional growth can come from use of the technologies in an open architecture that preserves the freedom to innovate. If we adopt the mind set that commoditized telecom and IT is basic infrastructure, we can struggle to keep the infrastructure open. In doing so we shall also keep open a seedbed of new economic development and new creativity and technology. In addition to a foundation for new business, this course will also support opportunity for further growth in customer support and education. But, if, as seems currently more likely, we follow the course of permitting the legacy industry a closed monopoly in order to save itself, we prolong the current agony and forego what economic development and growth could flourish in an open environment. The Commoditization of Everything We are witnessing the commoditization of the entire industry. It is not just telecom. It is telecom and all of information technologies. Both industries are finally maturing across the board. While new products are appearing, they cost less and do more. They bring a different kind of economic value. We are no longer likely to see the creation of any new industry giant like a Cisco or Microsoft. Open in architecture and cheap to produce, the new products are staggering the industry precisely because they are one or more orders of magnitude less expensive than the closed and proprietary systems they replace. While most new products are still designed in North America, Japan, or Europe, the majority of their components are made and assembled in Asia. There RAM is a commodity endlessly replicated in multi billion dollar "fabs." Around the 'corner' in Taiwan and other areas, commodity disk drives are mass-produced. A terabyte in the pocket is not far off. Commodity open source Linux and open source web services stand ready to do battle against Microsoft's closed systems. When new software is needed, it may be designed in North America or Europe. But the code is written in Bangalore, or Moscow, or Shanghai. Hua wei is sued by Cisco for doing what is in effect a commodity knock off. Back 'home" a handful of folk do the integration, first of the software, and then of the firmware and prototype hardware. They ship the result back to Bangkok or Kuala Lumpur for replication and assembly. Container ships bring the boxes back to ports like Yokahama, Newark, or Antwerp for sale on the shelves of Best Buy and Comp USA and other warehouse retailers. Prices are driven inexorably downward. This new, cheap and powerful hardware and software is being installed in networks with fiber cores terminating in monopoly controlled copper local loops. If the network is the Stupid network, in other words if it is an open access end-to-end Internet, boxes running commodity Ethernet can switch and route packetized information as appropriate. With the addition of a VoIP gateway card, the same devices can achieve vast cost savings by turning voice telephony into an application that rides alongside other digitized and commoditized applications such as email, web services, and television. But most networks are not open Internets. They are copper based networks with last miles subject to a monopoly controlled, content centric approach. They are bastions of legacy technology using an infrastructure far more complex and 10 to 100 times more expensive than that of the Stupid network. The legacy telco network is one where the monopoly must cut its own throat to try to compete with open architecture Internet upstarts that would take away its more profitable business customers. In other words, while the prices it receives for its services plummet, it must write down the value of its plant and equipment to a level where, given its smaller income, it cannot maintain its current cost and employment structure and sustain the ability to pay its debt. When voice no longer rides on the TDM transport that was especially designed to carry it and is just a packet-encapsulated application on an IP network, the new central office is no longer a building housing five million dollars worth of equipment. It fits on a desktop using SIP, SIP proxy servers, and ENUM databases. It costs well under five thousand dollars and delivers an entire range of services not possible to derive from now obsolete TDM hardware costing a thousand times more. Commoditization Makes More Job Losses Inevitable If telecom in the United States has lost 500,000 jobs in the past three years, with the inevitable demise of the LECs, it will lose another 500,000. Attrition in computer hardware and software should cause 500,000 more positions to evaporate. What is left will be administrative, financial product planning and marketing. Unfortunately, in this brave new world the marketers will be figuring out how sell $100 products in Best Buys rather than $100,000 systems to enterprises. As a result of this upheaval, customers will be left even more on their own. They will need help to figure out how to put the products together and assess what combination of products most effectively meets their needs. Until everything is truly plug-and-play and automatically-configured when attached to the network, education of the customer on product capability and system integration is the remaining critical area of competition. It is also the only bright spot for future industry employment. Commoditization dictates competition. But now that the companies are in trouble, competition on the part of the legacy, monopoly-owned, circuit-switched side of the telecom business is being allowed to disappear. Until the legacy companies go bankrupt and swap out their obsolete equipment, there can be no benefit to anyone from commoditization. The innovation and cost performance benefits of commoditization are all on the side of the open access, end-to-end, packet-based inter-networks. So far such networks in competition with each other for market share cannot make a profit. The mind-set of the political and regulatory system cannot comprehend the resulting paradox where the most productive and advanced networks cannot make money because, founded on commodity technology, they can be cheaply cloned with cookie cutter reliability. Staring into the headlights of the onrushing train wreck, it is blinded by fear of the destruction of shareholder equity and putting people out of work. It is seduced by the complaints of the incumbents who are selling the false premise: "Give us monopoly and we will have incentive to build." Determined to protect legacy interests, it consequently tilts the playing field in the US against the commodity players and against innovation. The truth is that, even with monopoly, will they not build. They will instead die, unless somehow, they managed to get use of the packet-switched, commodity-based technology successfully outlawed. To our great misfortune we do not yet understand that Commoditization has turned telecom and information technology into a basic enabling infrastructure like the electric, the water, the sewer and the highway grids. This new commoditized, powerful, and cheap technology can be used to deliver value and new jobs through preserving for everyone the ability to tinker and to innovate. Commoditization has removed the incentive for value creation from the monopoly network and left it with only the incentive to squeeze every penny of return for as long as it can prevent encroachment. The Legacy networks can and do use the new IP, commodity technology. But they are prevented by their debt obligations from being able to acquire enough of it. Furthermore, even if they could implement it extensively, their business model assumes a monopoly over access to transport. Because they must do it all, they are paradoxically denied its fullest advantage. The find themselves with no choice but to use packet switching and VoIP in an effort to sustain their traditional way of doing things. Protecting what they have, they lose what gains the new inexpensive equipment could offer. For example they would seek to offer international VoIP over their own dedicated network while a new competitor can dispense with the sunk cost of maintaining a physical network by simply renting access to an Internet that others maintain. Since the competition has only to rent access to transport and run voice as an application on that transport, it can offer service that is unencumbered by legacy costs. The business model of control of both applications and customers prevents productive investment. The paradox of commoditization leaves us with our uninhibited creativity as our only new source of economic development and growth. If all we do is drive prices down while maintaining the old structures, all we do is drive more people out of work. Failure to understand this leaves the legacy networks in power and able to preserve their obsolete assets by killing creativity, innovation, and experimentation. The backyard tinkerer has for the past century been the principal source of wealth in the United States. In the inexorable transition to a new commodity-based world, given current policy, we are exporting the freedom to be a tinkerer to Canada, Sweden, Japan and Korea. Having to compete in a global economy, we are sacrificing the viability of our resulting economic infrastructure in a foolish attempt to shore up legacy networks that can no longer serve as adequate means of competition. These are the lessons we carried home from Spring 2003 Voice on the Network where most were upbeat and pitching in to spread the new application April 7, 2003 -- ============================================================= The COOK Report on Internet, 431 Greenway Ave, Ewing, NJ 08618 USA 609 882-2572 (PSTN) 703 738-6031 (Vonage) cook(a)cookreport.com Subscription info & prices at http://cookreport.com/subscriptions.shtml The Paradox of Commoditization - Studies in VoIP and Telecom Economics, April -June 2003, 128 pages available at http://cookreport.com/12.01-02.shtml and http://cookreport.com/12.03.shtml =============================================================

4 3

June NANOG registration/topics
by Susan Harris 10 Apr '03

10 Apr '03

Registration is now open for NANOG 28, to be held June 1-3 in Salt Lake City. Our hosts are the Utah Education Network and the University of Utah: http://www.nanog.org/mtg-0306/ NEW at this meeting - sponsors can now display equipment during continental breakfasts and afternoon breaks: http://www.nanog.org/vendor.html Some of the topics we've lined up so far are: -- Cable backbone deployment issues -- Forwarding plane deadness detection -- Interaction between mobile and IP networks: operational issues -- Tracking lame DNS delegations -- Dual-stack deployment on the Abilene backbone -- Applications of the BGP Attack Tree -- BOFs: ISP Security, Device Discovery Protocols Send us your ideas for other talks - we're especially seeking proposals on: -- Provider case studies on network planning/design, protocol implementation, provisioning, useful tools, traffic engineering, troubleshooting, problems solved, & DoS -- Research on routing, network performance, statistical measurement/ analysis, and protocol development and implementation for operator feedback (10-minute slots) Call for Presentations: http://www.nanog.org/mtg-0306/call28.html

1 0

Salt Lake NANOG peering
by Pete Kruckenberg 09 Apr '03

09 Apr '03

We're excited to host the June NANOG in Salt Lake, and give the NANOG crowd to enjoy a few beautiful days in Utah. I know how important it'll be to have good connectivity to your network while you're here in Utah. As a special offer to NANOG attendees, we have made an exception to our normally stringent peering policy to ensure your SSH and IRC sessions are optimized for NANOG (Ethereal, ettercap and Cain&Able performance will not affected). For a limited time, we are now accepting requests for peering at PAIX Palo Alto and Equinix Chicago. As an additional incentive, anyone who peers with us before May 1 will be able to keep that peering indefinitely, so respond quickly. Pete.

4 3

TWTelecom
by Peder Angvall 09 Apr '03

09 Apr '03

Anybody from TWTelecom on the list who might know of a problem in Dallas? > 60% of pings appear to be lost through their Dallas network. I've been on hold at the NOC for 20 minutes. Packets: Sent = 908, Received = 275, Lost = 633 (69% loss), Approximate round trip times in milli-seconds: Minimum = 40ms, Maximum = 361ms, Average = 14ms 7 0.so-1-3-0.XL2.DFW9.ALTER.NET (152.63.99.214) [AS 701] 40 msec 40 msec 36 msec 8 POS7-0.GW10.DFW9.ALTER.NET (152.63.101.133) [AS 701] 40 msec 40 msec 36 msec 9 twtelecom-gw.customer.alter.net (65.208.11.194) [AS 701] 40 msec 40 msec 36 msec 10 core-01-ge-0-2-0-0.dlfw.twtelecom.net (168.215.54.85) [AS 4323] 40 msec * 40 msec 11 * * * 12 * * * 13 * * * 14 * 64-132-228-250.gen.twtelecom.net (64.132.228.250) [AS 4323] 44 msec 44 msec 15 64-132-228-6.gen.twtelecom.net (64.132.228.6) [AS 4323] 52 msec 48 msec 48 msec

2 1