- Test - lists.nanog.org

Re: [YA] Fwd: Class B Purchase
by Hank Nussbacher 07 Oct '98

07 Oct '98

At 10:48 PM 10/6/98 -0400, Jared Mauch wrote: >On Tue, Oct 06, 1998 at 07:56:59PM -0400, Kim Hubbard wrote: >> Michael, >> >> Could you define "legitimately multihomed" please? > > I think that a prime example would be a site such as >progressive networks, cnn (turner), msbnc, or any other high >traffic site. > > They will be multiply connected to major providers, >but most use some sort of ip director, and all their >machines sit on the same /24. > > Although I don't like the way some of these peoples networks >are built, they don't have enough machines/servers to justify anything >more than a /24 or even at most a /22, but still need provider >independent space as they may show up at regional exchange points. Been there. Done that. See: http://www.isoc.org.il/ipnum.html Requires me to scan Mae-East 4x a year and check whether they are still multihomed. Can such a method scale? Don't know. -Hank > > The best bet for these people in some cases is to go hunt for >space that is "available", as ARIN and folks won't allocate such space. > > Their other option is to sit on the network block of >one of their providers, but some people do have issues with >announcing other providers address space (like it's against their >internal policies). > > This would be a semi-justfiable use of a portable block >less than a /19, would it not? > > - Jared > >

2 1

Re: MAE NY?
by Aleksi.Suhonen＠saunalahti.fi 07 Oct '98

07 Oct '98

} Looks like a MFS MAE wins for me. Alex, you seem to be by an airconditioner } outlet in Newyork - Can we put a duct between our racks ;-) You're routing hot air now? ;-) } R -- Aleksi Suhonen Network Development Saunalahden Serveri Oy AXU-RIPE

1 0

Re: [YA] Fwd: Class B Purchase
by Hank Nussbacher 07 Oct '98

07 Oct '98

At 03:41 PM 10/6/98 -0700, David R. Conrad wrote: >Michael, > >>My advice in this instance would be to sue Sprint for antitrust violations >>because if you win then you get triple damages awarded and Sprint >>definitely has the bank account to pay out on the award. Forget the >>registries. They are just trying to do the best they can in the awkward >>situation that was created by Sprint. > >Wow. You're _really_ confused. > >The situation wasn't created by Sprint, it was created by the lack of self >control on the part of the other ISPs. Remember when the filters were >first instituted. Remember the growth of the routing tables. Remember the >maximum routing load the routers back then could handle. > >The registries _relied_ on Sprint's filters to give some teeth to "it's a >real good idea to go to your upstream". Sprint (read: Sean Doran) was the >ONLY isp to have the cajones to risk outrage to try to limit the >proliferation of long prefixes. If it wasn't for Sprint's filters, there AGIS used to have a page that stated the same routing filter policy as Sprint's. -Hank >would have been only registry whining as back pressure limiting the >allocation of provider independent prefixes. Geuss what: whining wasn't >particularly effective. RIPE-NCC and APNIC instituted fees for cost >recovery, and these had the side effect to limit organizations approaching >those registries for resources. InterNIC (at the time) was not able to >follow suit. Given the IAB's RFC 1814, there was _very_ little that >discouraged every Tom, Dick, and Mary consisting of two modems in a dorm >room that called themselves an "ISP" (or not) from demanding address space >from InterNIC. With the unilateral imposition of filters by Sprint there >was concrete evidence that prefixes longer than /19 were perhaps not a good >idea, thereby encouraging folk to go to their upstreams thus limiting the >proliferation of long prefixes. > >Yeah, let's sue. > >Frankly, with comments like this, I feel Sprint is approaching terminal >stupidity for keeping the filters in place. I'm sure they have a lovely >business case for keeping the filters active (or they wouldn't have lasted >this long), but at some point, natural selection has to be allowed to >function. I also think the registries should actually be registries and >not try to be the Internet's mommy. Internet service providers (or those >who think they are Internet service providers) ought to clean up their own >messes for a change instead of relying on Sprint and the registries to do >it for them. After all, everyone has a god-given inalienable right to >portable addresses, no? > >Disgustedly, >-drc > >

1 0

MAE NY?
by Brian Horvitz 07 Oct '98

07 Oct '98

Just curious to know who's exchanging traffic at MAE NY? If I am to believe what I see at www.mae.net, there's only two networks connected there, which would make it a pretty ineffective exchange point. With Pennsauken closed to new connections (anyone got extra rack space there?), I'd like to investigate an alternative for that area. Brian --- Brian Horvitz Shore.Net Network Engineering

10 16

Re: MAE NY?
by Richard Parker 07 Oct '98

07 Oct '98

On Tue, 6 Oct 1998, Jay Adelson wrote: > Too cold? Are you serious? Man, I just can't win. > > Our readouts were reporting 68-71F... Somebody tell me what this needs > to be set to for our exchange facilities. They're all pretty insanely > big, so this is something to do ONCE. Nows the time, we're deep into > engineering HVAC for the first fifteen. Oh, and I'd love to hear > humidity suggestions as well. Well, lets see... 1) Largest minimum operating RH for a communications device that I happen to have the specs for is 20%. 2) Smallest maximum operating RH for a communications device that I happen to have the specs for is 80%. 3) Static electricity is minimized when RH is at or above 35%. 4) RH below 25% can cause embrittlement of hygroscopic materials such as paper. 5) RH above 65% can cause mold growth and metal corrosion. 6) Humans are most comfortable when the RH is between 20% and 60%. 7) RH above 50% in cold weather can cause condensation inside of outer walls. So... perhaps a target humidity somewhere between 35% and 45% would be ideal. -Richard

1 0

Re: The number of the AS
by shields＠crosslink.net 07 Oct '98

07 Oct '98

In article <19981005230954.B3137(a)cp.net>, "J.D. Falk" <jdfalk(a)cp.net> wrote: > Not to be confused with the AS that shall forever live in > infamy, 7007. Anybody still filtering that one? Also 3830 (when it was reused). -- Shields, CrossLink.

1 0

Co-Lo for a customer -- OFF TOPIC
by Derek Elder 07 Oct '98

07 Oct '98

I need to get co-lo pricing and specifications for a customer. I need co-location space (preferably from the same folks) in: NYC Chicago San Francisco Minimum requirement at each co-lo is OC-48 redundant LEC connectivity. Must have carrier class HVAC (CO quality) and must be able to scale to 5 or more racks in 12 months. When you contact me, I'll also need to know how much backup capacity you have and what the backing is (UPS type, diesel, etc.). Any other info that you can provide, i.e., access, average temp/humidity, etc. would be very helpful. Over and out. D E R E K E L D E R Senior Engineer, CCIE# 4048 __________________________________________ USWeb Corporation http://www.usweb.com/ Network Solutions Practice 733 Third Avenue 7th Floor New York, NY 10017 ph: 212 301 2200 fax: 212 301 2201 mailto:delder@usweb.com USWeb - A Strategic Partner for the Information Age

1 0

NDAs
by Jennifer DePalma 07 Oct '98

07 Oct '98

Since a thread related to nondisclosure agreements has been brought up and I have been mulling over this topic lately, I'd like to know your opinions on whether or not the practice of requiring NDAs will continue. Are free market pressures working against such a model? Can secretive pricing really be sustained over a long period of time? It makes me think of when market economies first began. A lot of haggling took place, and buyers didn't always know what prices others were buying at. Some might buy the chicken for a quart of milk, others might have to hand over a whole cow. As the market matured, fixed pricing became more the rule of thumb for efficiency and expediency reasons. Bargaining still takes place but mainly with small "mom-and-pop" sort of shops and in more informal markets. A haggled down discount at Barnes & Nobles seems inconceivable. (Of course, a customer can haggle down a car salesman--or at least think he is--and individual car lots might receive periodic discounts from the factories...) An effective marketing strategy for a competitor of WorldCom might be open pricing. Am I right? Wrong? I'd like to see pricing become open, but I don't want the government outlawing NDAs. Perhaps NDAs are a market experiment that won't hold up? Thanks, Jennifer At 10:41 AM 9/16/98 -0400, you wrote: >last i checked www.spicegirls.com wasn't on the exodus network. > >i think it's on netkonect.net > >whitehouse.com is on exodus though... ;) > >as per the contractual issues, all they told lowly engineers like myself >was that we could disclose publicly the fact: > >1. we have reached a solution with GTEi >2. it includes expanding ds3's with them >3. we're not buying transit or paying for peering >4. we're like friends and stuff now... ;) > >i'm sure when bowman gets up, he'll have something to say. he probably has >a LOT more insight into #3 > >-dhiraj >sr. systems engineer, exodus communications > > > >On Tue, 15 Sep 1998, Adam Rothschild wrote: > >> > > Hmm.. not to start another massive thread here, but I wonder if this means >> > > that Exodus is now paying for transit and/or peering? >> > >> > Disclosure of this information is probally violation of >> > the agreement, so unless you work for one of them, or know >> > someone who does, you won't find out. >> >> Indeed. Point well taken. >> >> Still, I have a funny feeling that GTEI/BBN is being compensated royally >> for this. This would have never happened otherwise, methinks. >> >> I doubt Mr. Curran is doing this simply to allow himself and his customers >> to continue accesssing quality Exodus-hosted content, such as the Spice >> Girls and whitehouse.com. Then again, I could be mistaken! :-) >> >> > > ***************************** After September 25, please respond to jdepalma(a)alumni.princeton.edu Jennifer A. DePalma 202.789.5226 202.842.3490 (fax) Research Analyst Telecommunications and Technology Studies Cato Institute 1000 Massachusetts Ave., NW Washington, DC 20001

3 2

Re: MAE NY?
by David Lesher 06 Oct '98

06 Oct '98

It sounds to me as if what's needed is MAE-Byrd_Station. A further advantage is the limited number of backhoes on the continent. -- A host is a host from coast to coast.................wb8foz(a)nrk.com & no one will talk to a host that's close........[v].(301) 56-LINUX Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433

2 1

MAE NY?
by David Graves 06 Oct '98

06 Oct '98

I know that people continually talk about no space at 60 Hudson -- but does anyone have a human contact of the folks who run 60 Hudson, where I could verify this? -- David Graves Senior Systems Engineer, NYC Covad Communications 1-215-753-8952 Home Office 1-215-748-8628 Home/Cell

1 0