- Test - lists.nanog.org

No. of routers carrying full routes?
by dlc＠silcom.com 21 Oct '96

21 Oct '96

Where can I find/how can one estimate the number of routers carrying the full Internet route tables? I need to convince someone that singly-homed customer route flaps/ withdrawals should *not* propagate beyond our AS. I've found some discussion of this in the July NANOG archives, and talk about cisco floating statics, etc... and that "one-way" traffic is insignificant. I haven't upgraded to 11.1 yet, but my network is simple enough [read: not redundant enough?] that all my customers are singly-connected; in that case, statics in my border routers pointing to my customer edge-router seems to make fast install moot for now. Finally, is there a *searchable* archive of NANOG anywhere? Thanks. -- David Carmean WB6YZM DC574 <dlc(a)silcom.com> System/Network Administration, Silicon Beach Communications Unsolicited commercial e-mail not accepted. Violators will be LARTed.

2 2

Re: GigaRouter (Was Re: Cisco as Big Brother))
by Alexis Rosen 21 Oct '96

21 Oct '96

Neil J. McRae writes: > Well if you're talking i386. For NetBSD/i386 work is being done to > make gzip'ed kernels boot at thi time. So the author told me. Thing is, he's as bad in some ways as I am. Maybe it'll be done in two weeks, maybe not. But whenever it's done, I'm sure it'll be great. BTW, I'm curious why some people don't seem to like the idea. At the speed of modern systems, I suspect that loading uncompressed off flopy isn't much faster (if at all) than loading compressed off floppy and decompressing. In fact it may be slower. I'd test it right now but my arms aren't 4 miles long, which they'd have to be to reach the closest floppy drive on one of my NetBSDs... Heh. I can just see it. gzip becomes a kernel function. Although that's not really a joke- compressing filesystems aren't far off that mark... /a

2 1

Re: GigaRouter Latency?
by Vadim Antonov 21 Oct '96

21 Oct '96

Deepak Jain <deepak(a)jain.com> wrote: >Latency in a router is only somewhat important for most network >applications but on some applications it has a very serious effect. If >anyone would like to flame me, please do it in private email and I am >sure I will find the time to get back to you :). I'd like to point out that 1 ns of latency equivalent to 1 ft of wire (more-less), so 1 ms of latency is 190 miles of wire. It is not a serious delay, when compared with typical thousand-mile backbone links. (1 ms is a _lot_ of time to route a packet). --vadim

2 1

Re: GigaRouter (Was Re: Cisco as Big Brother))
by Vadim Antonov 20 Oct '96

20 Oct '96

I beieve my troubles with cisco flash were due to the relatively frequent updates (once a week or so) and somewhat high ambient temperatures (there wasn't adequate air flow). Solid-state nonvolatile memories age rather eapidly in those conditions. --vadim

1 0

Re: GigaRouter (Was Re: Cisco as Big Brother))
by Rob Liebschutz 20 Oct '96

20 Oct '96

> Alexis Rosen <alexis(a)panix.com> wrote: > > >Of course I'm not saying that I *Want* to use an HD in this situation; flash > >is clearly a big win. > > In my experience flash is a lot _less_ reliable than HDs. Ciscos losing > flash contents aren't that uncommon. Is this a general statement about flash, or just about the flash in Ciscos? I'd find it hard to believe that well designed solid state devices could be lest reliable than a disk drive, especially since a large part of a disk drive is solid state as well. > > >But I don't see how using a floppy or ZIP improves > >wins. > > Floppy doesn't rotate until it's boot time; and in COs dust is not > a big problem. > > So, for a boot device floppy is probably the best (providing it's not > an el cheapo device). An additional benefit is that it's easy for > field personnel to handle. > > --vadim >

2 1

Re: GigaRouter (Was Re: Cisco as Big Brother))
by Robert Sanders 20 Oct '96

20 Oct '96

Vadim Antonov <avg(a)quake.net> writes: > In my experience flash is a lot _less_ reliable than HDs. Ciscos losing > flash contents aren't that uncommon. ... > So, for a boot device floppy is probably the best (providing it's not In my experience, in an equal amount of hours of running Netblazers and Ciscos, the routers' flash has never failed me while the Netblazers *all* eventually failed to boot from floppy. The environment wasn't especially dusty, but it wasn't a clean room. I'd hesitate to use any floppy-booted device intended for use out of arm's reach. -- Robert

1 0

Re: GigaRouter (Was Re: Cisco as Big Brother))
by Rob Liebschutz 20 Oct '96

20 Oct '96

> > From: Michael Dillon <michael(a)memra.com> > > On Sat, 19 Oct 1996, Robert E. Seastrom wrote: > > > With cold-convenient-swappable IDE drawers that let even a > > kindergartener swap out an IDE hard drive and high quality 100mb hard > > drives available for like $50 > > Where on earth do you find new drives this small??? > > http://www.corpsys.com > > But apparently with their current stock of new drives you'll have to > settle for IBM 3.5" 170mb drives for $99 each ($89 qty 10). > > Now, if you don't mind refurbs, there are $65/$55 80mb drives and > $69/$59 131mb drives. I've dealt with these folks before and they're > righteous. > > Oh well, the $50 was a little optimistic, or maybe not if you're > buying a bunch of 'em... > > ---Rob > > > I have a hard time calling IDE drives reliable. At one time I used them in firewalls, and every now and then, tripwire would report random changed bits on the disk, and then the next time I ran tripwire, the problem would just go away. Now I use low end SCSI drives for these kind of applications cause I prefer to have all my data parity checked. Rob

1 0

GigaRouter Latency?
by Deepak Jain 19 Oct '96

19 Oct '96

Now I know a lot of people are going to jump on me for raising the latency question here, and I don't want to inflate its importance at all. Latency in a router is only somewhat important for most network applications but on some applications it has a very serious effect. If anyone would like to flame me, please do it in private email and I am sure I will find the time to get back to you :). How does the latency through a Gigarouter compare to the latency of say a NetBSD or FreeBSD box with (say) FDDI interfaces. [Ignoring the fact that most BSD boxes are running on a single PCI bus]. Basically I am curious how much advantage/disadvantage one gets by going from a Cisco to a Gigarouter solution based on issues like latency. I am sure the difference we are talking about is very slight and <2ms total, I just want to get an idea and I don't trust most of the technical sales personnel I have talked to to get a straight answer. I am very familiar with the latency of BSD (typical) boxes and those of Cisco Routers. Thanks, -Deepak.

2 1

Re: Cisco as Big Brother (Was Re: Cisco's AIP vs HSSI)
by Kent W. England 19 Oct '96

19 Oct '96

At 01:29 AM 10/16/96 -0500, Alan Hannan wrote: > > Although, I know lots of smart smart smart people who haven't. > One of them is looking for a job, and he's not looking at cisco. > (not me...) > > -alan > > > Alan; Somebody has to stay out of the cisco orbit to give the others a chance to figure it out. :-) Seriously, I started out with Proteon p4200 routers and I was one of the ones to talk Wellfleet into building a router (me and a hundred others) when what they wanted to sell were bridges. cisco came in third for the NEARnet startup. One of the reasons that cisco has earned its place as Number One is that, to my knowledge, cisco has never terminally screwed a client. No other contender can make that claim with me. I've been screwed by all of them, except cisco. They don't always give me what I want when I want it, but they don't do fundamentally bad things like walking away from the ISP market as Proteon did or wiping out their upper management team as Wellfleet did from time to time. cisco is no longer easy to work with in many respects, there is so much management for a large company. But in the early days, cisco was one of the first companies to turn their customer support inside out on the Net. You could (still can) talk to anyone in the company about almost anything. Not as easy as it was, but it can happen. But you can't get one engineer to hack something into the router code for you on just his say-so anymore. But once you could. I dunno maybe you still can, but I think you have to have megabucks behind you to do it. --Kent

6 6

Re: The Cidr Report
by Hakan Hansson 19 Oct '96

19 Oct '96

At 12:53 1996-10-18 -0700, Tony Bates wrote: > > Tony Bates <tbates(a)cisco.com> writes: > * Even though I try to stay impartial, a couple of things are worthy of > * note here. AS455 (Telianet) may want to check a possible leak as an > * increase of 455 Classful routes seems a little hard to belive. Others >[..] >Oops - I meant AS3301 ...apologies... Tony, We will take care of this obvious leakage asap. Thanks, Hakan

1 0