Mapping the Internet and Intranets

1
(No Transcript)
2
Mapping the Internet and Intranets

• Hal Burch, Bill Cheswick
• ches_at_lumeta.com
• http//www.cheswick.com

3
Motivations

• Intranets are out of control
• Always have been
• Highlands day after scenario
• Panix DOS attacks
• a way to trace anonymous packets back!

• Internet tomography
• Curiosity about size and growth of the Internet
• Same tools are useful for understanding any large
  network, including intranets

4
Related Work

• See Martin Dodges cyber geography page
• MIDS - John Quarterman
• CAIDA - kc claffy
• Mercator
• Measuring ISP topologies with rocketfuel - 2002
•   Spring, Mahajan, Wetherall
• Enter internet map in your search engine

5
The Goals

• Long term reliable collection of Internet and
  Lucent connectivity information
• without annoying too many people
• Attempt some simple visualizations of the data

• movie of Internet growth!
• Develop tools to probe intranets
• Probe the distant corners of the Internet

6
Methods - data collection

• Single reliable host connected at the company
  perimeter
• Daily full scan of Lucent
• Daily partial scan of Internet, monthly full scan
• One line of text per network scanned
• Unix tools

7
Methods - network scanning

• Obtain master network list
• network lists from Merit, RIPE, APNIC, etc.
• BGP data or routing data from customers
• hand-assembled list of Yugoslavia/Bosnia
• Run a traceroute-style scan towards each network
• Stop on error, completion, no data
• Keep the natives happy

8
TTL probes

• Used by traceroute and other tools
• Probes toward each target network with increasing
  TTL
• Probes are ICMP, UDP, TCP to port 80, 25, 139,
  etc.
• Some people block UDP, others ICMP

9
TTL probes
Hop 3
Hop 1
Hop 2
Hop 4
Hop 3
10
Send a packet with a TTL of 1
Hop 3
Hop 1
Hop 2
Hop 4
Hop 3
11
and we get the death notice from the first hop
Hop 3
Hop 1
Hop 2
Hop 4
Hop 3
12
Send a packet with a TTL of 2
Hop 3
Hop 1
Hop 2
Hop 4
Hop 3
13
and so on
Hop 3
Hop 1
Hop 2
Hop 4
Hop 3
14
Advantages

• We dont need access (I.e. SNMP) to the routers
• Its very fast
• Standard Internet tool it doesnt break things
• Insignificant load on the routers
• Not likely to show up on IDS reports
• We can probe with many packet types

15
Limitations

• Outgoing paths only
• Level 3 (IP) only
• ATM networks appear as a single node
• This distorts graphical analysis
• Not all routers respond
• Many routers limited to one response per second

16
Limitations

• View is from scanning host only
• Takes a while to collect alternating paths
• Gentle mapping means missed endpoints
• Imputes non-existent links

17
The data can go either way
B
C
D
A
E
F
18
The data can go either way
B
C
D
A
E
F
19
But our test packets only go part of the way
B
C
D
A
E
F
20
We record the hop
B
C
D
A
E
F
21
The next probe happens to go the other way
B
C
D
A
E
F
22
and we record the other hop
B
C
D
A
E
F
23
Weve imputed a link that doesnt exist
B
C
D
A
E
F
24
Data collection complaints

• Australian parliament was the first to complain
• List of whiners (25 nets)
• Military noticed immediately
• Steve Northcutt
• arrangements/warnings to DISA and CERT
• These complaints are mostly a thing of the past
• Internet background radiation predominates

25
(No Transcript)
26
(No Transcript)
27
Visualization goals

• make a map
• show interesting features
• debug our database and collection methods
• hard to fold up
• geography doesnt matter
• use colors to show further meaning

28
(No Transcript)
29
(No Transcript)
30
Infovis state-of-the-art in 1998

• 800 nodes was a huge graph
• We had 100,000 nodes
• Use spring-force simulation with lots of
  empirical tweaks
• Each layout needed 20 hours of Pentium time

31
(No Transcript)
32
The Internet has a diameter of about 10,000
pookies
33
Visualization of the layout algorithm

• Laying out the Internet graph

34
(No Transcript)
35
Visualization of the layout algorithm

• Laying out an intranet

36
(No Transcript)
37
A simplified map

• Minimum distance spanning tree uses 80 of the
  data
• Much easier visualization
• Most of the links still valid
• Redundancy is in the middle

38
Colored by AS number
39
Map Coloring

• distance from test host
• IP address
• shows communities
• Geographical (by TLD)
• ISPs
• future
• timing, firewalls, LSRR blocks

40
Colored by IP address!
41
Colored by geography
42
Colored by ISP
43
Colored by distance from scanning host
44
US military reached by ICMP ping
45
US military networks reached by UDP
46
(No Transcript)
47
(No Transcript)
48
History of the Project

• Started in August 1998 at Bell Labs
• April-June 1999 Yugoslavia mapping
• July 2000 first customer intranet scanned
• Sept. 2000 spun off Lumeta from Lucent/Bell Labs
• June 2002 B round funding completed
• 2003 sales gt4MM

49
Backhoes/truck bombs/mayhem

• The former happens surprisingly often
• Almost daily on the network of one major ISP
• 9/11 took out a fair amount of connectivity

50
CIDR and IP Counts
51
Routers in New York Citymissing generator fuel
52
Internet before 9/11/2001
53
Internet after 9/11/2001
54
Yugoslavia

• An unclassified peek at a new battlefield

55
(No Transcript)
56
Un film par Steve Hollywood Branigan...
57
(No Transcript)
58
fin
59
Intranets the rest of the Internet
60
The Pretty Good Wall of China
61
(No Transcript)
62
(No Transcript)
63
(No Transcript)
64
(No Transcript)
65
(No Transcript)
66
(No Transcript)
67
This was Supposed To be a VPN
68
(No Transcript)
69
(No Transcript)
70
Anything large enough to be called an intranet
isout of control
71
Case studies corp. networksSome intranet
statistics
72
Leak Detection
Mapping host
mitt

• A sends packet to B, with spoofed return address
  of D
• If B can, it will reply to D with a response,
  possibly through a different interface

A
D
Internet
intranet
C
B
Test host
73
Outbound Leak Detection
Mapping host
mitt

• Packet must be crafted so the response wont be
  permitted through the firewall
• A variety of packet types and responses are used
• Either inside or outside address may be
  discovered
• Packet is labeled so we know where it came from

A
D
Internet
intranet
C
B
Test host
74
Inbound Leak Detection
Mapping host
mitt

• This direction is usually more important
• It all depends on the site policy
• so many leaks might be just fine.

A
D
Internet
intranet
C
B
Test host
75
Inbound Leak Detection
Mapping host
mitt
A
D
Internet
intranet
C
B
Test host
76
Existence proofs of intranet leaks the slammer
worm

• Its a pop-quiz on perimeter integrity
• The best run networks (e.g. spooks nets) do not
  get these plagues
• Internal hosts may be susceptible

77
Some Lumeta lessons

• Reporting is the really hard part
• Converting data to information
• Tell me how we compare to other clients
• Offering a service was good practice, for a while
• The clients want a device
• We have gt70 Fortune-200 companies and government
  agencies as clients
• Need-to-have vs. want-to-have

78
Honeyd network emulation

• Anti-hacking tools by Niels Provos at
  citi.umich.edu
• Can respond as one or more hosts
• I am configuring it to look like an entire
  clients network
• Useful for testing and debugging
• Product?

79
Some open questions
80
How do you analyze a large graph over time?

• Five years of Internet data, mostly unanalyzed
• Alternate paths to a target country
• Sample insight Poland was off the Internet
  yesterday
• Placement of monitoring tools?
• Compute a display differences between two complex
  graphs

81
Visualizations

• These graphs are too big for a piece of paper
• Various approaches available, but none really
  satisfactory
• Build visualization graph as the data comes in,
  and as the network evolves

82
(No Transcript)
83
Mapping the Internet and Intranets

• Hal Burch, Bill Cheswick
• ches_at_lumeta.com
• http//www.cheswick.com

84
Some Internet Mapping Innards and Lessons

• Bill Cheswick
• ches_at_lumeta.com
• http//www.cheswick.com

85
Some of the Original Unix philosophy

• A tool should do one job, and do it well
• Pipes let us build powerful systems by linking
  tools together
• Kernighan and Pike, The Unix Programming
  Environment

86
Pipes let you feed the output of one program
directly into the next

• The programs must be written with this in mind
• Simple, constant formats
• One line per item
• Usually white space separates fields
• No header or trailer lines, please

87
Bad multiple files on one line.Worse different
on pipes and terminal
88
Bad stupid header line
89
Bad header lines. (duplicates Unixs traceroute
mistake)
90
Pinglist traceroute according to Unix specs

• Each input line produces a packet
• Each returning packet produces a single output
  line
• Ident field allowed the calling program to keep
  track of packets
• Original design was to generate UDP packets with
  varying TTL fields
• No DNS lookup that interferes with timing, and
  we might not want to do it, and other programs
  could do it better anyway
• Do one thing, and do it well

91
Original pinglist sample
pinglist 1 207.95.23.14 1 1 12.31.52.10 died 5
207.95.23.14 40 5 207.95.23.14 exceeded
92
Original pinglist sample
pinglist 1 207.95.23.14 1 Ident 1 Target 207.
95.23.14 Packet type UDP, TTL1
93
Typical sample usage cscanping all the hosts on
a class C net
( for i in seq 0 255 do echo i 1.i
50 done sleep 1 for i in seq 0
255 do echo i 1.i 50 done sleep 2 )
pinglist sort -n -k1,2
94
Pinglist -gt netio

• More parameters
• Can send pings (and a lot of other stuff)

95
Pinglist written in 1989

• Target and packet type fields are much more
  complicated now
• Sample target modifications
• Source routing
• 209.123.16.98.1,204.178.16.6
• Tunnelling
• 209.123.16.98204.178.16.6
• Type fields for ICMP, UDP, TCP with options, DNS,
  SNMP
• Its too complicated now

96
New version coming

• Pinglist classic pipe
• SGML packet descriptor type
• Others as we need them
• IPv6

97
Newest version of cscan
( for i in seq 0 255 do echo i 1.i P
ICMP probe done sleep 1 for i in seq 0
255 do echo i 1.i 64 UDP probe done )
netio -l 2 -x sort -n -k1,2
98
Simple (and ugly) traceroute
for i in seq 0 30 do echo i 1 i done
netio l 2 sort n k1,1 awk print 1,
2
99
Product prototype

• Based on netio
• Pipe networks into netscan
• Netscan pipes specific probe packets for many
  networks into netio
• Returning packets are piped back to netscan
• Netscan writes completed network information to
  standard output
• All the concurrency and scanning logic is in
  netscan

100
135.104.0.0/16
1 207.95.23.14 1
To Internet
netio
netscan
1 12.31.52.10 died
135.104.0.0/16 Path.....
101
135.104.0.0/16
To Internet
netio
netscan
rate
135.104.0.0/16 Path.....
102
Prototyping and debugging is easier

• Packet i/o is in text, easily written to a file
  for debugging
• Printf is easier interface than binary
• Ident field made it easy to match probes with
  their responses
• Negligible performance hit these programs are
  kernel- and network-bound, not CPU hogs

103
What a full scan produces simple text files!
104
Raw path data
162.83.64.0/19 Probe20041028 Target20041028162
.83.64.5 Destination20041028162.83.64.5
Path20041028,somerset,ping65.198.68.33,157.130.9
5.173,152.63.18.162,152.63.19.33,152.63.21.17,152.
63.21.81,204.255.173.25,205.171.17.61,205.171.8.24
6,205.171.230.10,205.171.8.218,205.171.209.114,205
.171.251.22,208.46.127.254,130.81.10.89,130.81.12.
146,130.81.5.170,162.83.64.5R9,9,8,8,6,14,11,11,9
,9,20,15,21,21,18,41,35,35S879,881,883,886,887,89
0,892,894,896,898,901,903,907,910,913,918,922,927
T255,254,252,251,250,249,249,250,251,250,244,245,2
44,243,243,241,241,240I30836,0,0,0,0,0,0,0,63731,
7129,0,0,0,0,63638,0,60220,52101
105
Raw path data, sample line broken out by fields

• 162.83.64.0/19
• Probe20041028
• Target20041028162.83.64.5
• Destination20041028162.83.64.5
• Path20041028,somerset,ping65.198.68.33,157.130.9
  5.173,152.63.18.162,152.63.19.33,152.63.21.17,152.
  63.21.81,204.255.173.25,205.171.17.61,205.171.8.24
  6,205.171.230.10,205.171.8.218,205.171.209.114,205
  .171.251.22,208.46.127.254,130.81.10.89,130.81.12.
  146,130.81.5.170,162.83.64.5
• R9,9,8,8,6,14,11,11,9,9,20,15,21,21,18,41,35,35S8
  79,881,883,886,887,890,892,894,896,898,901,903,907
  ,910,913,918,922,927
• T255,254,252,251,250,249,249,250,251,250,244,245,2
  44,243,243,241,241,240
• I30836,0,0,0,0,0,0,0,63731,7129,0,0,0,0,63638,0,60
  220,52101

106
Raw label datailookup output
144.232.1.237 sl-gw20-ana-0-0-0.sprintlink.net
204.178.16.49 209.194.240.14
(dns1.xspedius.net) 204.178.16.49 199.2.212.24
5 ip245.212.2.199.stingcomm.net
204.178.16.6 195.33.160.4 (ns1.att.nl)
65.198.68.67 203.213.192.2 (ns1.apnic.net)
65.198.68.67 199.18.24.10 owu-atm1-0s1.columbus
.oar.net 204.178.16.6 66.46.50.164
(ns1.business.allstream.net)
65.198.68.67 62.154.37.70 ma-eb1.MA.DE.net.DTAG
.DE 65.198.68.67
107
cat gt/tmp/12 12.0.0.0/8 ATT fnin -n
/tmp/12 ltlabels wc -l 2418
108
netsize
cat "_at_" awk ' BEGIN sum 0 // next
comments ignored /\s/ next lines
with only whitespace ignored NF gt 1 n
split(1, fields, "/") if (n ! 2)
print "ignoring invalid line (" NR ") " 0 gt
"/dev/stderr" next sum 2
(32 - fields2) END printf "0.0f\n",
sum '
109
cat gt/tmp/12 12.0.0.0/8 ATT fnin -n
/tmp/12 ltlabels wc -l 2418 fnin -n
/tmp/12 ltlabels netsize 34590812
110
Quick prototype question

• How many Kuwaiti routers did the scan find?

111
sed 10q labels 4.78.164.2
(ns2.Level3.net) 65.198.68.67 144.232.1.237
sl-gw20-ana-0-0-0.sprintlink.net
204.178.16.49 209.194.240.14 (dns1.xspedius.net)
204.178.16.49 199.2.212.245
ip245.212.2.199.stingcomm.net
204.178.16.6 195.33.160.4 (ns1.att.nl)
65.198.68.67 203.213.192.2 (ns1.apnic.net)
65.198.68.67 199.18.24.10 owu-atm1-0s1.columbus
.oar.net 204.178.16.6 207.230.214.82
randolph-k12-wi-us.customer.centurytel.net
128.2.198.127 66.46.50.164 (ns1.business.allstr
eam.net) 65.198.68.67 62.154.37.70
ma-eb1.MA.DE.net.DTAG.DE 65.198.68.67
112
awk 'print 2' labels sl-gw20-ana-0-0-0.spri
ntlink.net (dns1.xspedius.net) ip245.212.2.199.sti
ngcomm.net (ns1.att.nl) (ns1.apnic.net) owu-atm1-0
s1.columbus.oar.net randolph-k12-wi-us.customer.ce
nturytel.net (ns1.business.allstream.net) ma-eb1.M
A.DE.net.DTAG.DE
113
awk 'print 2' labels tr 'A-Z' 'a-z' sed
10q (ns2.level3.net) sl-gw20-ana-0-0-0.sprintlink
.net (dns1.xspedius.net) ip245.212.2.199.stingcomm
.net (ns1.att.nl) (ns1.apnic.net) owu-atm1-0s1.col
umbus.oar.net randolph-k12-wi-us.customer.centuryt
el.net (ns1.business.allstream.net) ma-eb1.ma.de.n
et.dtag.de
114
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' sed 10q ns2.level3.net sl-gw20-ana-0-0-
0.sprintlink.net dns1.xspedius.net ip245.212.2.199
.stingcomm.net ns1.att.nl ns1.apnic.net owu-atm1-0
s1.columbus.oar.net randolph-k12-wi-us.customer.ce
nturytel.net ns1.business.allstream.net ma-eb1.ma.
de.net.dtag.de
115
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' grep '\.kw' sed 10q
116
Answer

• None. Why?
• Kuwait.edu ! .kw
• Networks are aggregated
• Question how about Iran?

117
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' grep '\.ir' sed 10q fe0-0.lyra.bb.ni
avaran.tehran.sinet.ir n2-r3-c7206.iranet.ir n2-r2
-c7206.iranet.ir n2-r4-c7513.iranet.ir ns.parscybe
rian.ir ns1.nic.ir fa5-1-1.ipm-gw.bb.niavaran.tehr
an.sinet.ir ns.parscyberian.ir ns.parscyberian.ir
router1.tse.or.ir
118
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' grep '\.ir' wc -l 37
119
Question

• How many top-level country codes did the scan
  encounter?

120
Extract the first ten TLDs
121
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'print NF' sed
10q net net net net nl net net net net de
122
Collect and sort the TLDs
123
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'print NF' sort uniq
-c sort -rn sed 20q 86739 net 27564 com 6212
servfail 4568 jp 4089 au 2294 de 2054 ca 1982
ru 1623 br 1510 kr 1381 ar 1348 mx 1320 edu 1241
timeout 1182 it 1050 pl 1031 ro 1005 cn 896
arpa 794 in
124
Just two-letter TLDs
125
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c sort
-rn grep -v '\' sed 20q 4568 jp 4089 au 2294
de 2054 ca 1982 ru 1623 br 1510 kr 1381 ar 1348
mx 1182 it 1050 pl 1031 ro 1005 cn 794 in 789
id 709 uk 670 fr 603 ua 600 th 593 se
126
Check the end of the list for weird entries
127
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c sort
-rn grep -v '\' tail 1 gs 1 gn 1
gh 1 gf 1 fm 1 dj 1 ck 1 bt 1
bd 1 ac
128
Count tlds
129
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c sort
-rn grep -v '\' wc -l 177
130
Use pr to lay out results in columns
131
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c sort
-rn grep -v '\' pr -8 -t 4568 jp 533 tw
161 sk 42 nc 18 kg 8 ao 3 to 1
ne 4089 au 479 at 149 lt 41 bm 17 ug
7 tz 3 tm 1 ms 2294 de 424 nl 129 pt
40 ma 17 ba 7 pa 3 om 1 mr 2054
ca 423 fi 112 ec 39 su 16 zw 7 aw
3 mu 1 mn 1982 ru 408 sg 95 lb 39
bo 16 li 6 vi 3 mg 1 jm 1623 br
394 za 93 pe 38 yu 16 gi 6 tv 3
fo 1 io 1510 kr 391 il 85 si 37 ir
16 bf 6 rw 3 cx 1 gs 1381 ar 390 cz
83 vn 35 ge 16 am 6 ng 3 an
1 gn 1348 mx 366 be 76 ee 34 ve 15 qa
6 ky 2 tg 1 gh 1182 it 338 dk 71
sa 33 pf 15 mc 6 dm 2 pg 1
gf 1050 pl 334 co 69 ae 32 uz 15 fj
6 as 2 nr 1 fm 1031 ro 328 tr 61 by
29 ie 14 tn 6 ad 2 mv 1 dj 1005
cn 324 cl 60 uy 27 mk 14 kh 5 vu
2 gl 1 ck 794 in 313 bg 59 tt 26
np 14 jo 5 mw 2 cd 1 bt 789 id
282 es 57 lu 25 cy 14 gy 5 ml 2
bj 1 bd 709 uk 260 gr 52 cc 24 lk
14 eg 5 bw 2 bh 1 ac 670 fr 259 my
51 kz 23 py 11 mz 5 bs 2 ai 603
ua 257 hu 51 ke 21 mt 11 ci 5 az
1 va 600 th 237 pk 51 hr 20 ag 10
na 4 sv 1 tp 593 se 234 no 47 ni
19 dz 10 bn 4 sb 1 tj 580 nz 189 hk
45 md 19 do 9 hn 4 lc 1 tc 564
us 174 ph 45 is 19 cr 8 ws 4 fk
1 sr 544 ch 171 lv 45 gt 18 nu 8
sz 3 zm 1 rc
132
TLDs listed alphabetically
133
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c sort
-k1,1rn -k2,2 grep -v '\' pr -8 -t 4568 jp
533 tw 161 sk 42 nc 18 nu 8 ws 3
cx 1 gf 4089 au 479 at 149 lt 41 bm
17 ba 7 aw 3 fo 1 gh 2294 de 424 nl
129 pt 40 ma 17 ug 7 pa 3 mg
1 gn 2054 ca 423 fi 112 ec 39 bo 16 am
7 tz 3 mu 1 gs 1982 ru 408 sg 95
lb 39 su 16 bf 6 ad 3 om 1
io 1623 br 394 za 93 pe 38 yu 16 gi
6 as 3 tm 1 jm 1510 kr 391 il 85 si
37 ir 16 li 6 dm 3 to 1 mn 1381
ar 390 cz 83 vn 35 ge 16 zw 6 ky
3 zm 1 mr 1348 mx 366 be 76 ee 34
ve 15 fj 6 ng 2 ai 1 ms 1182 it
338 dk 71 sa 33 pf 15 mc 6 rw 2
bh 1 ne 1050 pl 334 co 69 ae 32 uz
15 qa 6 tv 2 bj 1 rc 1031 ro 328 tr
61 by 29 ie 14 eg 6 vi 2 cd
1 sr 1005 cn 324 cl 60 uy 27 mk 14 gy
5 az 2 gl 1 tc 794 in 313 bg 59
tt 26 np 14 jo 5 bs 2 mv 1 tj
789 id 282 es 57 lu 25 cy 14 kh 5
bw 2 nr 1 tp 709 uk 260 gr 52 cc
24 lk 14 tn 5 ml 2 pg 1 va 670 fr
259 my 51 hr 23 py 11 ci 5 mw
2 tg 603 ua 257 hu 51 ke 21 mt 11 mz
5 vu 1 ac 600 th 237 pk 51 kz 20
ag 10 bn 4 fk 1 bd 593 se 234 no
47 ni 19 cr 10 na 4 lc 1 bt 580 nz
189 hk 45 gt 19 do 9 hn 4 sb
1 ck 564 us 174 ph 45 is 19 dz 8 ao
4 sv 1 dj 544 ch 171 lv 45 md 18
kg 8 sz 3 an 1 fm
134
TLDs sorted by frequency, subsorted alphabetically
135
awk 'print 2' labels tr 'A-Z' 'a-z' tr
-d '()' awk -v FS. 'NF gt 1 print NF' grep
'..' grep 'a-z' sort uniq -c awk
'print 2, 1' sort grep -v '\' pr -8
-t ac 1 bo 39 do 19 hn 9 lk 24
ng 6 sa 71 ua 603 ad 6 br 1623 dz 19
hr 51 lt 149 ni 47 sb 4 ug 17 ae 69
bs 5 ec 112 hu 257 lu 57 nl 424
se 593 uk 709 ag 20 bt 1 ee 76 id 789
lv 171 no 234 sg 408 us 564 ai 2 bw 5
eg 14 ie 29 ma 40 np 26 si 85
uy 60 am 16 by 61 es 282 il 391 mc 15
nr 2 sk 161 uz 32 an 3 ca 2054 fi 423
in 794 md 45 nu 18 sr 1 va 1 ao 8
cc 52 fj 15 io 1 mg 3 nz 580 su
39 ve 34 ar 1381 cd 2 fk 4 ir 37
mk 27 om 3 sv 4 vi 6 as 6 ch 544
fm 1 is 45 ml 5 pa 7 sz 8 vn
83 at 479 ci 11 fo 3 it 1182 mn 1
pe 93 tc 1 vu 5 au 4089 ck 1 fr 670
jm 1 mr 1 pf 33 tg 2 ws 8 aw 7
cl 324 ge 35 jo 14 ms 1 pg 2 th
600 yu 38 az 5 cn 1005 gf 1 jp 4568
mt 21 ph 174 tj 1 za 394 ba 17 co 334
gh 1 ke 51 mu 3 pk 237 tm 3 zm
3 bd 1 cr 19 gi 16 kg 18 mv 2 pl
1050 tn 14 zw 16 be 366 cx 3 gl 2
kh 14 mw 5 pt 129 to 3 bf 16 cy 25
gn 1 kr 1510 mx 1348 py 23 tp 1 bg 313
cz 390 gr 260 ky 6 my 259 qa 15 tr
328 bh 2 de 2294 gs 1 kz 51 mz 11
rc 1 tt 59 bj 2 dj 1 gt 45 lb 95
na 10 ro 1031 tv 6 bm 41 dk 338 gy 14
lc 4 nc 42 ru 1982 tw 533 bn 10 dm 6
hk 189 li 16 ne 1 rw 6 tz 7
136
Data ! information

• The data collection is conceptually easy
• There are a lot of details that werent obvious
  when we started
• Packet rates over dial-up connections, where data
  speeds change dynamically
• How do you manage packet timeouts in such an
  environment?

137
A prototype for the venture capitalists
cat ltlt!EOF gtindex.html lthtmlgt ltbodygt lth1gtSummary
of Paths from Somersetlt/h1gt showsum
somerset lt/bodygt lt/htmlgt !EOF
138
Converting the data into information has been the
hardest part

• We are selling the third rewrite of the report
  now
• Difficult combination of data display, user
  interface, and network technical expertise
• Helpful to read Jef Raskin, Don Norman, and of
  course, Tufte

139
Some Internet Mapping Innards and Lessons

• Bill Cheswick
• ches_at_lumeta.com
• http//www.cheswick.com