A network by any other name....

[mangosteen]

Yesterday's phrase of the day was "professional pride".


I have a small obsession with naming things. I always have. The name is to control, and control is useful.

When I started at my current employer, the network naming scheme was a mess, mostly due to lack of consistency. You had to either grovel through DNS to find what was the most likely name, or go through the ARP tables to find the (most likely) MAC address of the switch hooked up to the router, etc. It was a nuisnace to do certain operational tasks, and as is the nature of things, operational tasks that would ordinarily be routine fell by the wayside. My obsession with naming things (in the guise of me) decided that this would not do.

So, not quite two years ago, I took lead on designing a network device naming convention that was cool enough to merit a paper at a networking conference. Some people loved it, a whole bunch hated it, and some really wanted to talk with me about it, but in general, people appreciated that it was written, because it meant that someone put some serious thought into the topic, specifically in the realm of network devices.

Sometime before publication, my colleagues and I implemented it network-wide. Doing a traceroute into our network is actually somewhat aesthetically pleasing, now. You can re-derive every network link, have an understanding of where the packets are going, and it all basically makes sense. As an example, traceroute to lychee.net.tufts.edu (130.64.2.99). If you look at the last couple of hops near the end, you can see names of the pattern "A-x-B". That indicates an internal transit link. The other end of the connection is labeled "B-x-A". There's also parts of the spec or terminus interfaces, and single-address-multiple-interface devices, like switches. It's all named, logical, and descriptive. It pleases me.

I told you all of that, so I could tell you this:

For all that I attended Columbia University In The City Of New York, Inc. for 4.5 years, I never had any interest in how their university network was put together. This has since changed. So, in the process of doing a traceroute to the one server I knew still existed (i.e. www.cs.columbia.edu), I noticed the following.

[~]> traceroute www.cs.columbia.edu
traceroute to shadow.cs.columbia.edu (128.59.23.100), 64 hops max, 44 byte packets
 1  tab-rtr-pri-v-vlan1-primary.net.tufts.edu (130.64.1.2)  0.398 ms  0.389 ms  0.306 ms
 2  core-pri-x-tab-rtr-pri.net.tufts.edu (130.64.4.98)  0.300 ms  0.251 ms  0.308 ms
 3  border-pri-x-core-pri.net.tufts.edu (130.64.4.9)  0.695 ms  1.043 ms  0.696 ms
 4  nox230gw1-Vl-506-NoX-TUFTS.nox.org (192.5.89.21)  1.874 ms  1.378 ms  1.480 ms
 5  nox230gw1-PEER-NoX-NOX-192-5-89-10.nox.org (192.5.89.10)  7.145 ms  6.896 ms  7.153 ms
 6  nyc-gsr-abilene-nycm.nysernet.net (199.109.4.129)  6.739 ms  6.681 ms  6.915 ms
 7  columbia.nyc-gsr.nysernet.net (199.109.4.14)  6.947 ms  6.539 ms  6.613 ms
 8  cc-core-1-x-nyser32-gw-1.net.columbia.edu (128.59.255.5)  7.533 ms  7.084 ms  6.706 ms
 9  mudd-edge-1-x-cc-core-1.net.columbia.edu (128.59.255.86)  7.322 ms  7.162 ms  6.754 ms
10  shadow.cs.columbia.edu (128.59.23.100)  7.133 ms  7.054 ms  7.538 ms

Note the bolded bits. Sure enough, if you make guesses about the address for the other end of hop #9, you notice the following:

[~]> host 128.59.255.86
86.255.59.128.IN-ADDR.ARPA domain name pointer mudd-edge-1-x-cc-core-1.net.columbia.edu
[~]> host 128.59.255.85
85.255.59.128.IN-ADDR.ARPA domain name pointer cc-core-1-x-mudd-edge-1.net.columbia.edu

So, someone at my alma mater thought that my ideas on network naming were cool enough to adopt for use in their own network. I was grinning like an idiot for several hours when I found that out. It isn't the first time that a 3rd party has used a tool/scheme/framework I have written, but it's nice to have something to which I can point and say "someone else thought this was a good idea, too!".


There are any number of reasons that my alma mater will likely never name a building after me. I don't have tens of millions of dollars to donate, and I can't conceive of a building ugly and useless enough to meet Columbia's criteria for new edifices, to name a couple.

.....but it appears I left my mark on the University, just the same.

Comments

[muffyjo]

W00t!

You are such a geek! (hehehe)

[arfur]

that was cool enough to merit a paper at a networking conference.

"merit." heheheh. "merit."

inanely yours,
-bb.

[also-huey.livejournal.com]

This reads like some part of a Big Dork Purity Test. I think anybody who reads it is a bit of a dork, anybody who understands it is a bit more of a dork, and anybody who says "Dude! Really cool!" is a dork dorkity dork dork.

Dork.

[mangosteen]

Dork.

[she fierces, wildly]

[lifecollage.livejournal.com]

FYI, it's a good thing I tend to call my dear spouse before beginning raging flame wars in his LJ. [fierces again, half-heartedly]

Hi there!

[also-huey.livejournal.com]

An old (like, twenty years ago) girlfriend pimped something called 'nonficwrimo' to me. I laughed and laughed and laughed.

[lifecollage.livejournal.com]

And now, having tracked back through the friends lists, so am I. Laughing, that is.

What is the sound of one world shrinking? zzzzip!

So, you joining?

[also-huey.livejournal.com]

A friend of mine is an auto mechanic. On his time off, he writes music.

I write nonfiction. On my time off, I fix lawnmowers. If you're having a 'mowerfixmo', I'm your guy, but I just spent eight hours writing about software that I'm rapidly coming to hate, and not planning on doing any more writing until Monday morning.

[mama-pipa.livejournal.com]

Guess I am a dork too, because I thought that was cool!

[debkitty.livejournal.com]

Understood? Only the part where you merited an entire conference, the result of which was to one-up our alma mater (look what I came up with that you didn't). Kudos (although I prefer full fledged chocolate)!

[fieldsnyc.livejournal.com]

Sweet.

[docorion.livejournal.com]

Well, I didn't understand it completely, but I'm enough of a dork to think it's pretty cool. Well done!

[slinkr.livejournal.com]

Dude! That's really cool.

[awfief.livejournal.com]

quite a week, between that and the Chinese!

[mangosteen]

Indeed.
It appears that I'm getting what I wished for, and getting it good and hard. :)

[awfief.livejournal.com]

in the butt!

sorry, couldn't resist.

[cos]

Ahah, so that's what you were presenting at NANOG that time. Nice.

[whitebird.livejournal.com]

Oh, I like that a lot. Too bad our networking people are dumb as rocks and would never implement something as spiffy.

Is the paper readable somewhere ... ooo. Paper, and video stream. Cool!

[mangosteen]

Ya know? I never actually saw the video stream before.

On one hand, "Wow. That's a lot of 'well's and 'um's and 'say's."

OTOH, I actually sound genuinely competent, and slightly authoritative. I had no idea. :)

[whitebird.livejournal.com]

I saw it after I left the comment. It was a pretty good presentation.

On the one hand, yes, yes you do. It's a hard habit to break, but if you can, people pay a lot more attention to you. (And, I believe, their comprehension levels go up, too.)

On the other hand, yes, yes you do. I may show the presentation to one of our computing services wonks who actually has a brain and see if anything comes of it.

[debkitty.livejournal.com]

So where's the link to this video stream? I haven't seen you in a while, I haven't heard your voice in a while, and I need my fix. :)