July 1996 - NANOG - lists.nanog.org

Re: T3 or not to T3
by sob＠academ.com 22 Jul '96

22 Jul '96

> I'm not clear on the distinction -- but since the first location we > want to do this would be based in San Francisco, I'm referring mostly to > mae-west, the pacbell nap, and CIX. Only one of these is an NSF NAP (PAC*BELL in San Francisco). > >Generally for each connection to each provider, you would have to set up > >BGP. > > Yeah, definately. But most backbones seem to have "customer routes" as > an option, and if I trust them enough to get those routes correct then > I will hopefully not have to bother with extreme amounts of filtering. If you pursue getting that option, there will still be some routes that you will get via a transit connection that will have to pass through some interconnect. That means you are still going to to depend on some interconnect somewhere. Of course, if you don't have a transit agreement anywhere, you just won't see these routes at all. > It's pretty easy to enforce "no transit" at the packet filtering level > -- only packets destined for my nets will be allowed in. Is there some > other aspect of filtering I'm forgetting about? We have a dedicated > and backup network engineer at any rate. The border router would be a > cisco 7200 or 7500 series with 128Mb. The main thing is to insure that you don't get route announcements from places you don't expect them and that you do get them from places you do expect them. Expectations will change from time to time as changes in the routing mesh do occur. That means that your routing folks will have to monitor for such changes and make adjustments from time to time. The RA (and others) have been doing such monitoring and you can get some idea how such things affect you by looking at some of their work. -- Stan | Academ Consulting Services |internet: sob(a)academ.com Olan | For more info on academ, see this |uucp: {mcsun|amdahl}!academ!sob Barber | URL- http://www.academ.com/academ |Opinions expressed are only mine.

1 0

Re: sell shell accounts?
by Brett D. Watson 20 Jul '96

20 Jul '96

> > Acceptable arguments are: > > > o Switches can handle more throughput > > That's difficult to quantify in theory *or* practice. > > > o Switched networks are easier for humans (or at least, humans without > > huge brians && many internal registers) to design/debug/maintain. > > More levels of indirection does not mean it's easier for humans necessarily. > In fact, there are many more nobs to miss, and more places for error to > be introduced into your engineering model. i would agree strongly. putting a bunch of switches between two routers makes troubleshooting that much more involved. if i can't get a packet from router A to router F (assuming B,C,D,and E are atm switches) where has the link broken down? you have to go to each switch and look at incoming/outgoing cells to see which is not sending or receiving. this scenario definately makes more work for the network operator doing the troubleshooting. it certainly doesn't make things *easier*. -brett

1 0

Re: sell shell accounts?
by Brett D. Watson 20 Jul '96

20 Jul '96

> On Fri, 19 Jul 1996, Avi Freedman wrote: > > > I assume that the mondo-routers have a switching engine, and the > > only job of the main CPU/gated combo is to feed the current idea > > of the best next-hop/interface to the engine. > > Current mondo-routers like Netstar uses a switch fabric to parallelise > switching. It becomes pretty obvious that at OC-N speeds current serial > backplane architecture doesn't scale > > Handling routing updates, unlike switching packets, is more a function of CPU > and software routines. and the last time i worked on a netstar gigarouter (admittedly months ago) it was a long way from really doing solid routing (bgp, etc). maybe it's much improved now? -brett

1 0

RE: sell shell accounts?
by David Whipple 20 Jul '96

20 Jul '96

What BGP issues are you talking about? As I understand it, The Netstar gigarouter simply runs gated (sort of). So you have the standard issues beween gated and Cisco, but that's it. Thanks. David Whipple. >---------- >From: Karl Denninger, MCSNet[SMTP:karl@mcs.com] >Sent: Friday, July 19, 1996 2:34 PM >To: freedman(a)netaxs.com >Cc: vansax(a)atmnet.net; richards(a)netrex.com; agislist(a)interstice.com; >nanog(a)merit.edu >Subject: Re: sell shell accounts? > >> Nope, remember - there is no magic. Any mesh of PVCs that one makes >> over a switched network must reflect the toplogy of that network, and >> one can set up a matching set of active routing sessions and route >> weights which will cause traffic to flow the same way. >> >> Yes, the switches are a bit faster and have less to do. Data moves through >> them in a few ms less per point. But as you said (and as I said in our >> discussion in NYC), relative to any distance, the speed of light guarantees >> that you won't notice the difference. >> >> The question is: Will there be routers available that can make IP >> routing decisions based on 40-60kroutes and move 2-3 OC3s worth of >> bidirectional traffic? The building of the configs to have a routed >> network work the same as a switched ATM one can be automated, but it's >> true that it *can be* easier to see what's going on in a large-scale >> switched network. >> >> Avi > >Uh, yes. Like today. > >See NetStar (recently acquired by ASCEND). > >Their hardware does nothing *slower* than DS-3 right now, and updates the >route forwarding table 20x/second. :-) > >This thing is a monster, and will make the so-called "state of the art" >7513s look like dogmeat when in full production (presuming the BGP issues >get completely resolved). > >Coming soon to a network near you. > >-- >-- >Karl Denninger (karl(a)MCS.Net)| MCSNet - The Finest Internet Connectivity >http://www.mcs.net/~karl | T1 from $600 monthly; speeds to DS-3 available > | 23 Chicagoland Prefixes, 13 ISDN, much more >Voice: [+1 312 803-MCS1] | Email to "info(a)mcs.net" WWW: >http://www.mcs.net/ >Fax: [+1 312 248-9865] | Home of Chicago's only FULL Clarinet feed! >

2 1

IS IT REALLY FULLY MESHED (Was Re: sell shell accounts?)
by Mike Trest 19 Jul '96

19 Jul '96

The point of discussion orginated as a question about a fully meshed network in AGIS. IMHO. When all you do is connect ATM PVCs point-to-point between routers you do not, even remotely, have a fully meshed ATM cloud. If AGIS onstructed this way [*Please I am not saying they are*], then there is only a series of routers which use ATM in between for (a) go fast point-to-point or (b) lower price-per-megabit point-to-point (c) all of the above. The topology would look like this. And it is exactly as Avi indicates a perfect match for the router-only network topology. R1 --(ATM PVC)-- R2 --(ATM PVC)-- R3 If however, it is truly a cloud with routers at the edges, the topology can be described like this: R1 --(PVC R1,R2)--( ATM CLOUD )--(PVC R1,R2)-- R2 R1 --(PVC R1,R3)--( ATM CLOUD )--(PVC R3,R2)-- R2 ( ATM CLOUD )--(PVC R1,R3)-- R3 ( ATM CLOUD )--(PVC R2,R3)-- R3 Using this approach, then the ATM interface card in each router has a path to all other routers, then it is fully meshed. HOWEVER, THIS ASSUMES that (1) a series of ATM switches is used to create a physical cloud capable of touching all the routers. (2) That the PVCs are provisioned accordingly. I have observed a growing number of network providers calling themselves "...ATM based..." who are only purchasing point-to-point PVC service from a carrier. Until you purchase a PVC from each router to all others in your network, it is not fully meshed. Even then, it is the routers which are managing the mesh unless you also own the ATM switches in between. ..mike.. Mike Trest, ATMNET Voice: 619 643-1805 5440 Morehouse Drive Fax: 619 643-1801 San Diego, CA USA 92121 Pager: 619 960-9070

1 0

Re: Why aren't ISPs providing stratum 1 NTP service?
by Sean Donelan 19 Jul '96

19 Jul '96

>So the list works... if you don't want to provide "public" services, adjust >the server to allow connections only from your own IP blocks. Currently I, and the company I work for, don't mind providing some services to the "public" Internet community. The company almost always gets more out of it than it costs to supply. I just don't want to get trapped into "supporting" public services. I get enough hate mail now when one of our no charge services goes out of service on occasion. A related, but more on topic for NANOG, issue is aligning more network services with network topology. I don't think putting a NTP stratum 1 server on the NAP network fabric is a good idea. I do think providers exchanging NTP across the NAP fabric is a good idea. Redundant voting catches a bunch of dumb errors. A packet saved, is a packet you don't have to carry. Its not as huge a problem as multiple MBONE tunnels transiting the same physical lines. But I'd rather have multiple associations at the edge of my network, and pass NTP in a structured manner around inside my network. The same thing is true for several other network services. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

2 1

Anybody want to face Bob Metcalfe on TV?
by Michael Dillon 19 Jul '96

19 Jul '96

If anyone would like to face Bob Metcalfe on TV here is the place and time. Of course I don't know if there will be audience questions allowed so don't take my word for it, check into it on your own. ---------- Forwarded message ---------- Date: Thu, 18 Jul 1996 18:35:55 -0600 From: Jack Rickard <jackr(a)boardwatch.com> Reply-To: inet-access(a)earth.com To: "'inet-access(a)earth.com'" <inet-access(a)earth.com> Subject: Re: ONE ISPCON SCHEDULE FINALIZED Resent-Date: Thu, 18 Jul 1996 20:14:31 -0600 (MDT) Resent-From: inet-access(a)earth.com We've had a somewhat interesting development. Apparently C-SPAN has taken a bit of interest in this gathering of Internet Service Providers. Dr. Robert Metcalfe is doing a session Saturday afternoon - essentially summarizing what he sees at the show and covering his IMMINENT COLLAPSE OF THE INTERNET theme. It would appear C-SPAN is proposing to televise this session. Peter Granoff is one of a handful of Grand Sommelier's (I can't spell that - wine guys) in the country and is based in San Francisco. He has been selling wine on a web site for over 18 months now. We had asked him to come do a realistic session on marketing specialty products on the Internet describing the realities. He was interviewed on KCBS in San Francisco this afternoon about his session. I understand we appeared in Microtimes today. John Quarterman, author of the first Internet book THE MATRIX will be honored at the John C. Dvorak Awards banquet Saturday night along with Tim Berners-Lee. How am I doin' Justin? Jack Rickard ===================================================== Jack Rickard Boardwatch Magazine Editor/Publisher 8500 West Bowles Ave., Ste 210 jack.rickard(a)boardwatch.com Littleton, CO 80123 303.973.6038 voice 303.973.3731 fax http://www.boardwatch.com ======================================================== ============================== ISP Mailing List ============================== Email ``unsubscribe'' to inet-access-request(a)earth.com to be removed. Do not post flames to the list -- if you must flame, use private email.

1 0

Re: Why aren't ISPs providing stratum 1 NTP service?
by rrv＠uiuc.edu 18 Jul '96

18 Jul '96

At 9:37 PM 7/17/96, Sean Donelan wrote: >Or if the NTP folks aren't >thrilled with the prospect of 7,000 different ISP's wanting to be listed, >perhaps a note a the top of the list directing people to contact their >ISP first. That sounds like a good idea!

1 0

Re: problem at mae-west tonight?
by matthew＠scruz.net 18 Jul '96

18 Jul '96

Here's an edited copy of mail I just sent elsewhere, which I believe deserves some thought by other network operators : Just to update you on this situation: The problem was caused by a bad NetEdge device on the link from Netcom to MAE-West. It caused a partial failure of layer 2 connectivity from July 13 08:30 PDT until July 15 16:00 PDT. That's over 2 days. The current design and implementation of the route server suffers from two fatal flaws which, until they are fixed, will mean that I avoid using route servers at all costs: 1. The route server does not have access to the true state of layer 2 interconnectivity, and thus cannot adjust its announcements to avoid "black holes" in the layer 2 fabric, be they caused by missing ATM PVCs or bad bridging devices. Clearly we need a protocol whereby the route server can gain information from the routers about who they can and cannot see. (Others here have suggested an approach similar to the new OSPF multipoint model) 2. The route server does not have a mechanism whereby a client router can force the route server to NOT advertise information to another client, except by editing the policy information and waiting for it to reload. That time-frame is excessive from a network operations standpoint, because every minute that the route server announces a route into a black hole, customers are without service... service which can be restored simply by tearing down the session with the route server, which of course forces it to stop announcing routes to _every_ client... solving the one problem, and causing others at the same time. -matthew kaufman matthew(a)scruz.net ps. Given that I'd like to refuse to use the route servers, should I _not_ peer with them, but make my RADB policy look right, so that other people can tell what my policy is? or should I peer with them, to help their statistics gathering, but set my policy to "don't advertise anything to anyone" ?

2 1

RE: The root nameservers will be replaced August 1st
by Jim Fleming 18 Jul '96

18 Jul '96

On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: <snip> @ @ More interestingly, if someone wants to create an alternative set of @ root servers, there is no particularly good reason for them to be located @ at exchange points, (unless I am confused about what networks are all @ about...). True, servers at exchange points should exhibit greater @ availability, but that is probably not the largest challenge faced by @ alternative root servers. @ These are very good points. I think that the exchange points are just convienant co-location sites. Also, in the future, there may be other services on those boxes that have not been announced yet. It is useful to have the boxes in strategic locations in advance of additional changes. -- Jim Fleming UNETY Systems, Inc. Naperville, IL e-mail: JimFleming(a)unety.net

4 6