- Test - lists.nanog.org

User-level ISP performance measurement
by John R Levine 01 Dec '97

01 Dec '97

I hope I'm not reopening a huge can of worms here, but ... Some of my relatives have started a small ISP in western Vermont. (They already run the local telco.) They're wondering how to compare the performance that their users see to the ISPs in the adjacent town. I told them that there isn't anything published that'll be of much use, since the stuff I've seen tries to measure backbone performance and for the most part does so ineptly. For user level stuff, it seems to me that you could pick a representative set of actions (visit large and small web sites, download e-mail, scan newsgroups), dial in via PPP every hour for a week being sure to flush out caches and otherwise start from a consistent user state, do the actions, time them all, and make some simple stats. This would be miserable to do manually, but pretty straightforward to do automatically. Is there anything available, either as a product or service, that does this sort of analysis? One of the other hats I wear is as a director of a software testing tools company, and we could certainly whip up this kind of stuff in no time using our existing products if there were a demand for it. Regards, John Levine, johnl(a)iecc.com, Primary Perpetrator of "The Internet for Dummies", Information Superhighwayman wanna-be, http://iecc.com/johnl, Sewer Commissioner Finger for PGP key, f'print = 3A 5B D0 3F D9 A0 6A A4 2D AC 1E 9E A6 36 A3 47

1 0

Re: The Year of the Backhoe Continues...
by Sean Donelan 01 Dec '97

01 Dec '97

>do people really think that the relative fiber cut rate is any higher than >in the past? there's a bit more fiber, hence the term 'relative'. but i >suspect that the predominant factor is that new victims have a better means >of broadcasting their displeasure. Your tax-dollar at work, I wrote the FCC and asked them this morning. Here is the reply a I got back (very impressive response time, less than 2 hours): "For the five consecutive quarters prior to the Third quarter of 1996, cable outages were higher than they have been since then. Nevertheless there continues to be s significant overall increase in the number of cable outages since the start of outage reporting in the third quarter of 1992. Typically, about half of the cable outages are dig-ups. This information is is an informal estimate and its accuracy is not guaranteed or in any way certified by the FCC. The outage reports we receive are only available in hard copy. The FCC outage database is a rough internal working tool and not suitable for formal distribution, therfore it is not made public." So I think that means the answer to Randy's question is no. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Status: RE: The Year of the Backhoe Continues...
by Bownes, Robert M. (EXCH) 01 Dec '97

01 Dec '97

Wiltel's cable was cut by a front-loader that was in a disposal yard digging outside of their right-of-way (where they should not have been digging). Wiltel is currently locating their hand-holes. They are stretching out a temporary patch cable using 3M mechanical splices to attempt temporary restoration. No ETR at this time. >---------- >From: Nathan Stratton[SMTP:nathan@robotics.net] >Sent: Monday, December 01, 1997 11:52 AM >To: Bownes, Robert M. (EXCH) >Cc: 'nanog(a)merit.edu' >Subject: Re: The Year of the Backhoe Continues... > >On Mon, 1 Dec 1997, Bownes, Robert M. (EXCH) wrote: > >> Just got a note from our NOC... >> >> > >> >Worldcom fiber failure in Dover FL....techs enroute.... >> > >> >249 ds-3's down....ETA for Worldcom tech 10-15 mins >> > >> no ETR > >Is this due to the "fiber upgrade"? > >

1 0

The Year of the Backhoe Continues...
by Bownes, Robert M. (EXCH) 01 Dec '97

01 Dec '97

Just got a note from our NOC... > >Worldcom fiber failure in Dover FL....techs enroute.... > >249 ds-3's down....ETA for Worldcom tech 10-15 mins > no ETR

2 1

RADB
by Abha Ahuja 01 Dec '97

01 Dec '97

Hi All! Just so you all know, DNS for ra.net has been fixed. Thank you, Bill Manning! -abha ;) __________________________________________________________________________ -------------------------------------------------------------------------- abha ahuja ahuja(a)merit.edu Merit Network, Inc. 313.764.0294

1 0

RE: Eugene Kashpureff's Defense Fund
by Bob Allisat 01 Dec '97

01 Dec '97

>1. Is this an organization for Domain Names for registering them >and maintaining them , and developing global policy ?? CORE is just a private company attempting to market five domain names. .WEB and .ART do not belong to CORE because other companies have developed registries for these Top Level Domains. However CORE is trying in the process to independantly establish Global regulations over the namespace and this is highly unacceptable to most reasonable observers. They've as much right to register .FIRM, .STORE and the others as Internic has to do the same for .COM, .ORG and .NET and Alternic to act as a registry for .XXX, .ZOO, .WEB and so on. None of them have the right to govern the Internet. >2. If this Person has breached federal telecommunication regulations ? >if so. then must pay for the crimes as set down in the countries law.!!!! Mr. Kashpureff is innocent until proven otherwise. And by all appearances the charges are all politically motivated and not very well grounded in legalities. No-one actually owns the domain names. So what the fellow did is not illegal in the strict sense of the word. And it really didn't inconvenience anyone except for those who wish to claim ownership over a universal resource. >3. Defense Fund , if people want to help this person , this should not >be connected in anyway to this organization or this may be seen around >the world as an incorrect organization policy !!!!!!! I think it would be wonderful if CORE were to donate some of it's money to the Eugene Kashpureff's Defense Fund! I also think it would be great if they officially endorsed the Eugene Kashpureff's Defense Fund. Unfortunately CORE is itself trying to steal Internic's monopoly and install one of it's own. And CORE will in all probablility be the next target of protest actions. >4. I am confused over the role of the CORE know because of this discussion >on a issue that doesn't concern the domain registration or GLTD policy >development. This whole debate is central to the debate. Internic holds a monopoly over Domain Names. No-one else can participate at any level. Now CORE wants to take over as the Monopoly. Leaving everyone else out in the cold. And by everyone else I mean individuals, small companies, non- profit organizatrions and most nations in the so-called "third" world. The price tag for a new domain name registry is a half million US dollars (US$500,000). Not only that CORE is a private organization run without any public oversight under Swiss law. It is, furthermore, *not* supported or sanctioned by any member State of the United Nations nor by the UN itself. CORE is a rogue DNS elephant like Internic was in the past. We may have no alternative but to stop it in any way we can. And Eugene points the way very clearly. Will CORE stop and reverse it's course? That remains to be seen. If not expect more of the same protest actions like Kashpureff's. We will have no other choice. TeleVirtually Yours, Bob Allisat Director, World TeleVirtual Network http://www.wtv.net PO Box 191 St E Toronto Canada M6H 4E2 tor(a)wtv.net (416) 534-1999 http://www.wtv.net/portfolio.html

7 8

Re: Horrible Service Agreements
by Sean Donelan 30 Nov '97

30 Nov '97

Yet another person who wasn't going to say anything, but.... If you prefer not to read this, here is a cisco configuration which will fix it config t no ip routing For the clue-impaired, it is not advised you follow any of preceeding or succeeding advice without consultation with a competent advisor. >As for the equipment issue, I feel little sympathy. I have had to go in >and clean up swamp pits of archaic hardware and software. I am not saying >that's what you are, but I am saying that the service providers are quite >reasonable in setting minimum hardware standards for BGP customers. Remember I think the concern isn't over the service provider setting minimum standards for BGP customers, but the open-ended nature of the service provider unilaterally change their standards. Normally a material change to the service by the provider allows the customer to decide if they want to continue with the revised service, remain with the existing service, or end the service without penalty. >And finally. Life's too short to obsess about these kinds of issue. The time to worry about contracts is before there is a problem. Although there are a lot of badly written contacts, and I would walk away from them rather than trying to 'educate' the provider. In practice most of the terms exist to cover a problem that happened in the past. Telco affiliated ISP's use the boilerplate prohibiting you from using their name due to the number of long-distance telephone resellers that used to call people at dinner time and say they where calling on behalf of AT&T, but were actually a boiler room operation in Nevada whose only connection to AT&T was the logo on the telephone equipment they were using. Useful boilerplate language exists, see your lawyer: "The customer may advise end users that certain services are provided by the Company in connection with the service the customer furnishes to end users; however, the customer shall not represent that the Company jointly participates in the customer's services." Similar boilerplate exists for most of the things Forrest Christian is worried about. But dealing with lawyers at billion dollar companies is often an exercise in futility. Unfortunately most NSP contracts cover all the reasons why the NSP can terminate service without notice, but don't actually bind the NSP to providing an adequate level of service or any service in some cases. I would worry about ISPs with 'simple' contracts as much as ISPs with 'complex' contracts. Simple contracts are harder to write well. Most Acceptable Use Policies actually read like unacceptable use policies. The NSF AUP listed eight acceptable uses and only two unacceptable uses. >If you run your ISP right, these won't be a problem. If you don't run >your ISP right, these won't be your real problem. In the time you have >been upset about this, you could have a well written AUP with procedures >in place that would make any transit provider happy. You are making a big assumption. If it would have made the transit provider happy, why didn't the transit provider contract list this as an acceptable option? Even if the salesperson says its ok, six months later after the salesperson is long gone, and there is a problem will there be any record the transit provider found the ISP's AUP or ISP's procedures acceptable. Or will they apply the one strike and you are out rule, and terminate service without notice. Unless it is included within the four corners of the contract, I would not make any assumptions about what would or would not keep any transit provider happy over the lifespan of the contract. Although, if the transit provider is unhappy, I suspect they will find some reason to terminate the service irregardless. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Cat5500/Core6000/?
by Brian Horvitz 30 Nov '97

30 Nov '97

Anybody have real comparison experience between the Catalyst 5000/5500 series switches and the 3Com CoreBuilder 6000? Seems to me that the core builder is going to be able to push around a lot more packets. I'm just really looking for likes and dislikes on either or, if anybody just hates them both and has anything else they like. Brian Horvitz

1 0

wait before you flame
by Nicholas Bastin 30 Nov '97

30 Nov '97

I did not intend that last message to start a whole gigantic thread about ipv6/ipv8, I just want to know if anyone has drafted anything about ipv8... Just to head off the onslaught. Nick

4 4

ipv8
by Nicholas Bastin 30 Nov '97

30 Nov '97

This may or may not be the place to ask, but...Can anybody point me at some technical documentation/proposals on ipv8? Nick

1 0