- Test - lists.nanog.org

RE: multi-homing fixes
by smd＠clock.org 28 Aug '01

28 Aug '01

"David Schwartz" <davids(a)webmaster.com> writes: | I'm not sure I believe that this tragedy of the commons exists where people | route on allocation boundaries. If I make Sprint carry an extra route just | for my little network, that helps all Sprint customers reach my little | network. Cool, so you can get them to contribute a fraction of the 50 cents you owe me? Since I live downstream from Sprint, I tell you what: I'll give you the first flap for free. And that's a generous offer, given that your little network represents approximately zero percent of my traffic, and I doubt I'll be complaining to my other personalities that I can't get at your fascinating content without subsidizing your network's globally-visible dynamicism. Sean. (who does, incidentally, subscribe to some pay-for-play web sites, for example, which helps them pay for a part of THEIR dynamicism. i must remember to ask for a 50 cent * n discount at renewal time...) P.S.: the point here, since it's easy to miss, is that a clearing house function is useful for containing complexity of negotiation, and is currently being done by the RIRs on behalf of their owners business opportunity: a tool which reliably lets one decide if prefix x is worth carrying in all its flapping glory, or whether it should be bitbucketed until a bill is paid

2 1

wow, this is off-topic but very interesting (nanog bounce messages)
by smd＠clock.org 28 Aug '01

28 Aug '01

Wow, it seems that I'm a male chauvinist pig. Sean. | From: System Attendant <MAIL-SA(a)syndesis.com> | To: "'smd(a)clock.org'" <smd(a)clock.org> | Subject: ScanMail Message: To Sender, sensitive content found and action t | aken. | MIME-Version: 1.0 | | Trend SMEX Content Filter has detected sensitive content. | | Place = davids(a)webmaster.com; patrick(a)cybernothing.org; randy(a)psg.com; | rmeyer(a)mhsc.com; nanog(a)merit.edu; | Sender = smd(a)clock.org | Subject = Roeland Meyer owes me money! (was re: multi-homing fixes) | Delivery Time = August 28, 2001 (Tuesday) 10:16:36 | Policy = Sexual Discrimination | Action on this mail = Quarantine message | | Warning message from administrator: | Sender, Content filter has detected a sensitive e-mail. |

2 1

RE: multi-homing fixes
by smd＠clock.org 28 Aug '01

28 Aug '01

AVG writes: | Therefore, the cost of extra prefix is definitely not $0.50; it is _much_ | higher. Quite likely, considered globally. But personally, I'll be happy to get Roeland Meyer's prompt payment of the 50 cents (per flap, per router, per prefix) he owes me, as a start in discovering a market price for my routing slots. I urge the rest of you to invoice him for your 50-cents-times-n, today. Sean.

2 1

RE: multi-homing fixes
by smd＠clock.org 28 Aug '01

28 Aug '01

| ARIN could decide to issue /32's to dialup customers so they could | change providers without renumbering and it would do no good since nobody | would carry those routes. This is an excellent suggestion, and I have made it to RIPE people from time to time when this issue comes up. Some slow-start fine-tuning means that technonerds with DSL, cable, dialup and so on and several PeeCees might grow into a /28 someday! (Or is that a /48?) | [It would] harm the net as a whole if RIRs adopted a | microallocation policy that ... resulted in ... allocating | non-routable IP space. Not really. It would force some issues that people would rather not deal with. Most notably, it would force the issue of how providers exchange reachability information, which has languished while the same providers have focused on how providers exchange traffic attracted by that reachability info. Right now, the RIRs seem to understand how the conversations along the lines of "if you send me <prefix x> i will route it to you as long as you route <prefix y> to me" enough that they don't actually have to happen bilaterally. Maybe that could change, given that at least one of the RIRs is developing some opinions on how those conversations SHOULD work rather than how they WOULD work. Any catharsis of spurious morality couched in "for the good of the Internet" (or the small guy or the small ISP or the big banks, or whoever your favourite cause is) rhetoric is always welcome, by me. | Don't make any sense. That was my point in replying to you. There is no | right to a route in my router. If you want a route in my router, you better | find out what routes I'm willing to carry and under what terms. Is it still, loosely, "did you get this prefix/mask allocated by an RIR"? Or are things breaking down already? Sean.

1 0

RE: multi-homing fixes
by Roeland Meyer 28 Aug '01

28 Aug '01

|> From: David Schwartz [mailto:davids@webmaster.com] |> Sent: Monday, August 27, 2001 8:52 PM |> |> Patrick Greenwell wrote: |> |> > Please explain why the "basement dual-homer" should not |> have the same |> > right to diversity as the "major services." |> > |> > And please, be specific. |> |> The quesetion is bogus, there is no such thing as a |> right to have a route |> in my router without paying me for it. If I am paying for those routes then I have a contractual right. If you don't want my redundant feed, and the route advertisements that go with it, then don't take my money. No one is advocating multi-homing without payment. Where did you read that? |> If I choose to extend |> that privilege |> to people who meet certain minimum requirements because I believe the |> benefits will outweight the costs, then that's *my* right. Please detail the exact costs of a, BGP inserted, routing table entry. Is it, maybe, 50 cents? Now, how much are you getting for a DS1 link? What does that cost, exactly, considering that an outfit capable of setting up multi-homing are probably the folks that your techs never hear from, but once a year? That appears to be a margin that is far above keystone. How greedy do you want to be? |> All others can |> pay me to do it if they want me to. Your rights end at my network. BTW, randy's position is rather strange, coming from someone that used to support the FidoNet community, by being the FTSC chair.

7 6

RE: multi-homing fixes
by smd＠clock.org 28 Aug '01

28 Aug '01

Patrick Greenwell asks: | The question is would you honor microallocations from a RIR if they | said "we have designated this particular space for microallocations, would | you please accept routes for these netblocks with these prefixes?" This is a business opportunity for the ex-MERIT folks who maintained the PRDB in the days of the weekly updates. Resurrecting and "improving" as-out:-style, less-dynamic approaches to partition recovery will earn you zillions! Sean.

1 0

Roeland Meyer owes me money! (was re: multi-homing fixes)
by smd＠clock.org 28 Aug '01

28 Aug '01

Roeland Meyer <rmeyer(a)mhsc.com> writes: | Please detail the exact costs of a, BGP inserted, routing table entry. Is | it, maybe, 50 cents? Deal. Kindly send 50 cents (in U.S. currency) for each prefix you announce to the world, since I am carrying full routing on one of my own personal routers. You can get my postal address by email. My employers no doubt would like their 50 cents too, however I expect there is a possibility that they will interpret your sensible offer to be on a PER-ROUTER basis. Please let me know where they should send the invoice for the few hundred dollars per prefix. Don't forget that when you change your NLRI at or near the source you will be occupying a different slot, so it's only fair that you should send 50 cents again to me (and to my employers * number_of_routers) when that happens. Of course, we could just burn the existing NLRI into something approaching ROM, which will save you some recurring fees, during those pesky flapping periods of advertisments and withdrawals, but perhaps the dynamicism is worth it to you. Let me know. It's pretty easy to implement this in the vein of http://www.ripe.net/docs/ripe-210.html After the useful historical demonstration of a "if you flap three times, go home for the rest of the day" policy, damping more aggressively along the lines of "if you flap three times, you're out of my tables forever" is quite feasible. And, if you're a net.genius with lots of luck, maybe this is exactly the sort of 50-cent-times-n savings that appeals to you! Sean. - -- Sean Doran <smd(a)ab.use.net> Evil Greedy Bastard

1 0

OC3 Transit Service in New York
by Elijah Kagan 28 Aug '01

28 Aug '01

Please excuse me if this is not a right list to ask such questions, but... I am a network manager for NetVision, an Israeli ISP. We have a couple of locations in US which primary purpose is getting transit services from US ISPs. One of these ISPs is Teleglobe and we have an OC3 transit connection at 60 Hudson, New York. For the past couple of months we are growing discontent with service levels received from them and are looking for an alternative provider to work with. ISPs considered so far are C&W, Global Grossing and Qwest. Could someone comment on their network services, customer service, technical skills of their staff, etc. Any other real life experience with the above ISPs is welcome. Especially I am interested in information about Global Grossing because of their very attractive proposal. (I did not mention UUNet because we are already connected to them). Recommendations for other ISPs are welcome. Also, I would like to know about other people's experience with Teleglobe. Thanks, Elijah P.S. I know that NANOG stands for North American, but can someone recommend an ISP in Brussels, Belgium (except UUNet) to buy a transit OC3 from? Thanks again.

4 4

Measuring PoP to PoP latency--tools to use?
by Murphy, Brennan 28 Aug '01

28 Aug '01

What tools exist that will periodically log into multiple routers and run ping tests to various destinations and then record the results in graphical form? I am trying to avoid writing an expect script to do this. I want to set up something that records semi-live data to produce a web page similar to: http://ipnetwork.bgtmo.ip.att.net/current_network_performance.shtml The map portion of the graphic may be too fancy. A simple grid similar to this might be better: http://www.internethealthreport.com Any suggestions as to what tools to investigate would be greatly appreciated. I browsed the nanog archives as well as google...no hits so far. Thanks, BM

7 6

RE: multi-homing fixes
by Roeland Meyer 28 Aug '01

28 Aug '01

|> From: Randy Bush [mailto:randy@psg.com] |> Sent: Monday, August 27, 2001 7:51 PM |> |> the point of 2xDS3 was specifically to get major services, |> and not to get |> every basement dual-homer. do the latter and you have the |> same grazing |> of the commons as we have today. Considering the cost of a DS1, in California anyways, I might disagree with you. DS1 is steep enough, DS3's are much too steep and sounds much too anti-competitive. Randy, you also know the rates the rest of the planet pays for E1's and E2's, if they can even get anything that fast. If you want a speed-bump then don't build a speed-farging-mountain. What you are saying is that businesses that can't afford the monthlies of a pair of DS3's aren't allowed to be in business? Is that an official position? How would you like it if you were in that position? Can I quote you on that?

4 4