- Test - lists.nanog.org

CIX Filtering...
by salo＠msc.edu 14 Oct '96

14 Oct '96

Does anyone remember when the CIX proposed to filter networks of customers of CIX members which were not themselves CIX members? Was it as early as 1995? (We are trying to review a paper on Internet exchanges. It seems ironic that this effort by the CIX to restrict traffic was rendered ineffective by the same technique the CIX used to overcome the impediments created by the NSFNET AUP, namely alternative paths for "unacceptable" traffic.) Thanks, -tjs

5 7

Re: generators, etc ...
by Sean Donelan 14 Oct '96

14 Oct '96

In article <m0vCc8h-001K9SC(a)on-ramp.ior.com>, davids(a)on-ramp.ior.COM (David J. Schmidt) writes: > An editorial in Unix Review (October issue I believe) talked about > some outages that their site had suffered and mentioned other > outages. Then the writer asked the question "Who's at fault?". His > answer was that the consumer was at fault, because the consumer is > unwilling to pay the rates necessary to pay for the level of service > that they demand. > > We all scream for fixes when these outages occur. How much are we > (and our customers) willing to pay for them? Heavy sigh. Even if you are willing to pay more you won't find anything better to buy. I've been trying to buy better service for years, and haven't found a seller willing and able to commit to deliver anything better. Multi-homing to multiple naps, leasing separate lines, maintaining a 24 hour NOC, and general management of the whole thing is expensive and time consuming. I would love to outsource the whole mess and go back to making money on our real product. There are customers out there willing to pay more, because they are already paying a lot more. They are fairly easy to spot because they are doing lots of crazy things like multi-homing, leasing their own lines, buying big cisco's and connecting them to NAPs. Show me you can deliver. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: BBN outage
by Apu 14 Oct '96

14 Oct '96

On Fri, 11 Oct 1996, Charles R. Hoynowski wrote: > We're trying to figure out how long BBN's been down, are you talking > 12:30am PDT? If so, then they been down 16 hours. I started seeing mail > failures to Stanford (a big BBN site) at 9:33am PDT and first got email > about the problem from another ISP at around 11am. I haven't seen anybody answer this, and while I suppose everybody could call the BBN NOC and ask, that's probably not a useful way for them to spend their time. Therefore, for those who don't know but want to, and for the record (archive)... According to BBN ticket # 60414, the problem started at 03:27 UTC (October 11, 1996). As of October 12, 1996 01:54 UTC (the latest entry on the ticket), BBN believes all affected customers should be back online. This is the last entry on the ticket as of now. _ | | |-| | |pu <apu(a)spfld.com>

1 0

Re: CIX Filtering...
by salo＠msc.edu 14 Oct '96

14 Oct '96

Thanks again, -tjs

1 0

Re: CIX Filtering...
by salo＠msc.edu 14 Oct '96

14 Oct '96

Thanks, -tjs

1 0

Re: Creating exchanges [Was: Re: MAE-East - 30%]
by Barry Raveendran Greene 14 Oct '96

14 Oct '96

Hello Robert, Robert E. Seastrom wrote: >Hong Kong looks slightly promising except for the slight problem of >next summer. Singapore is probably a no-go because of how anal the >government is there; all we need is an AP CIX with a host government >enforced AUP that says "no dirty gifs, no politically unacceptable >speech, etc.". Actually, when you count the number of countries directly interconnected at any Asia Pacific IX (including the ones on the US West Cost) STIX - Singapore Telecom Internet Exchange is currently #1. (see http://www.stix.net) Singapore's SBA (the Internet censor) has stated over and over again that business and international traffic are NOT their problem. They just out to protect the people at home - and that's Singapore's really none of our business (and not a topic for NANOG). Please read the information from the source vs reading biased reporters from the US - http://www.gov.sg/sba. As to you other point, AP CIX is a no go in Asia Pacific. This discussion as been going on for over 4 years. I've been involved and actively trying to make it happen for over 3 years. The cable consortium, telco, national, AP ISP and US ISP politics prevent any type of successful 'neutral' IX in Asia Pacific. Barry -- Barry Raveendran Greene | || || | Senior Consulting Engineering | || || | Singapore | |||| |||| | tel: +65 738-5535 ext 235 | ..:||||||:..:||||||:.. | e-mail: bgreene(a)cisco.com | c i s c o S y s t e m s |

1 0

Re: generators etc...
by Zachary DeAquila 13 Oct '96

13 Oct '96

please read my note again. I was NOT flaming your statements, I agree with them 100%. I was flaming Mr. Michal Dillon's responses to your statements. Apologies if that was unclear. --Zachary On Saturday, Oct 12, 1996, Mike O'Dell writes: >please read my note again. > >I was NOT saying how I could have done better. >that was PRECISELY my point. >

3 2

Re: Intra/Inter - was Inet-II
by William Allen Simpson 13 Oct '96

13 Oct '96

> Date: Fri, 11 Oct 1996 22:36:26 -0700 > From: Vadim Antonov <avg(a)quake.net> > I'm sorry that this is getting more and more off-topic... > Yes, so I'll keep my reply short. And hope that the flaming anti-government yahoos stop filling my mailbox.... > William Allen Simpson <wsimpson(a)greendragon.com> wrote: > > >Delivery of all social services has clearly been shown to be cheaper > >when done by government, when costs of fund collection, advertisement, > >and administration are included. Private charity is always more > >expensive, even when partly done with volunteer labor. The best numbers > >I have at hand are 3% government (US Food Stamps) versus 17% private > >(United Way) for administrative costs to delivered services. And when > >advertising and fund collection are added, many non-profit "charities" > >spend upwards of 50% on overhead! > > Huh? Sure you forgot to figure in the cost of collection of taxes > and counted only disbursement (and did you count enforcement and > what groceries spend on processing food stamps?) > I explicitly (in my quoted comment) did _NOT_ forget the cost of fund raising, enforcement, or any other administrative cost. And neither did the United Way study, nor the several others that I remember reading.... This is _very_ well established, folks. The US IRS is a model of efficiency for fund raising, with far less distributed overhead than others, and a budget of 1/10,000 (off the top of my head) of collected revenue. Compare with sales or property tax, with their thousands of separate administrations, and own elected officials, etc. I have read that groceries actually _like_ to get food stamps. They put big signs on their windows "food stamps accepted here". They are much easier and cheaper to administer than manufacturers' coupons. (As an aside, related to my later recycling note, groceries _hate_ recycling, which would _not_ happen without government enforcement.) > Most of what charities spend is cost of fundraising. > That's correct, for many! > When you figure _that_ in, you'll get government efficiency well > below that of any private organization. > This comment makes no sense, as it is pretty obvious that government does not spend most of its budget on fund raising, and therefore your conclusion is at variance with your own assertions. You have obviously not looked at the numbers, nor been on the boards of any private non-profits, as I have. > And, of course, it is always much simplier to get funds for > "charity" by telling people they'd better pay or go to jail. > If it is much simpler, then it is likely cheaper, too. There is no way that private organizations can compete on an efficiency basis. > What you compare is charity and legalized racket. > > In fact, the social cost of welfare is enormous. It created > generations of people which cannot and don't want to earn living. > A culture of entitlement, if you like. No charity ever made > recipients "entitled" to anything. > This is a descent into diatribe, and demonstrably untrue.... Your biases are showing, and no rational discourse is likely. Sounds to me like you've been listening to political advertisements for your economics. Folks, I just gave a few of the well-documented examples of when it is more efficient for the people to form a collective organization (called government) than to privatize. Please keep your facts straight. > >We already had commercial telecommunications companies; they gave us > >X.25 ... and now, ATM. > > X.25 was fine for its time. It still processes all credit card and > ATM transactions. ATM (another one) is going to be dead pretty soon, > as new generation of IP routers will roll out. > The reason we chose X.25 for ATM transactions here in Michigan (X.25 was the 3rd protocol stack I ever wrote, circa '78-79) was because it was the only thing available! The alternative was direct leased lines, and we did a fair amount of that, too. But, it was _never_ "fine". I preferred IBM bisync and Burroughs poll-select, and even Honeywell VIP, and invented a few of my own. > And, ah, yes. ATM figured prominently in NAP-related papers, ok? > vBNS is ATM-based. _Who_ is promoting that insanity? > ATM figured prominently in _PRIVATE_ company responses to the NAP bid. ATM was proposed by a _PRIVATE_ company for the vBNS. I do not recall ATM being mentioned in the NSF RFP, and the private ATM-based NAPs had a _lot_ of problems coming up! Our Internet is a fine example of what government and academics _can_ do. And Internet II may very well be a boon to the world, if it can provide impetus for development of new technologies. If it just becomes another AUP vehicle, then that would be a loss. It is our job to make sure that it does not! WSimpson(a)UMich.edu Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32 BSimpson(a)MorningStar.com Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2

1 0

Re: Creating exchanges [Was: Re: MAE-East - 30%]
by Dave Curado 13 Oct '96

13 Oct '96

> Hong Kong looks slightly promising except for the slight problem of > next summer. Singapore is probably a no-go because of how anal the > government is there; all we need is an AP CIX with a host government > enforced AUP that says "no dirty gifs, no politicallly unacceptable > speech, etc". Of course it remains to be seen what will happen in HK next July, but note that HK has a reasonable sized nap already. The HKIX. My understanding is that they are about to upgrade the exchange to an ATM switch. (whatever) However, I had huge problems running an ISP in HK. The telco pricing has always been such that it is scads less expensive to connect back to the US than to another country, even another country just a few miles away. I keep watching the pricing and goings-on to see if/when there will be a reasonable pricing decrease on bandwidth to Asia, but haven't seen anything yet. I'd love to see someone get an oc3/12/48 pipe from HK to the west coast, and resell to the 60+ HK ISPs at reasonable prices. While in HK a week ago, I was trying to set up a 128K line for a small organization there. The cost is in the US$4k range -- I was afraid to tell those guys what I pay for a 128K bri line here in the US. For the time being, it seems that the most affordable solution would be for each country to build it's own exchange, and have everything link back via the US. (although its silly to have traffic from HK to Bangkok routing via San Jose...) If someone has more recent or more optimistic info, please let me know! davec

3 3

Re: generators, etc....
by Sean Donelan 13 Oct '96

13 Oct '96

>I suggest that neither gleefull hand-rubbing nor "obvious" pronouncements >based on partial knowledge about the real situation will seem quite so >appropriate when it's *your* turn to be downfield from the launcher. Don't forget 500 year floods that happen twice in a couple of years, e.g. the midwest. Or the occasional bombing, e.g. our New York City POP is in Hackensack, NJ now, the MIDNET router in Oklahoma City is in a landfill now. An Internet connection is pretty far down on the list of important things in a real disaster. Stuff happens. No one can predict or plan for everything. But you can have procedures in place to combat the biggest problem in diasasters, poor communication. Partial knowledge about the real situation happens when the real information isn't made available. Techies sometimes get too wrapped up in the hardware, redudant fiber, generators, bomb-resistant shelters; and forget about keeping people informed. It also works to your advantage to keep people informed. They tend to be more understanding, and willing to give you time when they know the situation, and that you are dealing with it. But you only get one free pass. Non-statement statements tend to fan the flames. I think I've said all this before. BBNplanet is better than many ISPs, once you find out the ticket number. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

2 1