IPv6 Matrix Exec Summary Dec 2011 Results - ICCA Pondicherry 31 Jan 2012

IPv6 Matrix Project

Tracking IPv6 connectivity Worldwide
http://www.ipv6matrix.org

Dr. Olivier MJ Crépin-Leblond – ocl@gih.com
December 2011 Update for ICCA January 2012

Page 1 IPv6 Matrix Project - http://www.ipv6matrix.org

We are running out of IP addresses

http://www.potaroo.net/tools/ipv4/index.html
When we reach this point, it will be too late since
there will be no more “free” IPv4 addresses!
Real time data collected September 2011

IPv6 Matrix Project
ISOC England was awarded a Community
Grants Programme award in November
2009
Design and implementation of an “IPv6
Crawler,” software on a computer that
crawls through the DNS at regular intervals
in order to detect and test:
IPv6 DNS servers
IPv6 compliant Web servers
IPv6 compliant SMTP mailers
IPv6 compliant NTP servers.


Project Rationale
Today, more than 95% of Internet traffic is generated
by a small number of data sources – i.e. the world’s
busiest Web Sites
Without IPv6 accessible content, IPv6 has no chance
of being used - ever.
Take the 1 Million most popular Web site list from
alexa.com as a starting point for the domains to be
tested. Add more domains later.
Test them for IPv6 connectivity
This is equivalent to testing about 6.3 million hosts
worldwide
Use GeoIP database to estimate real host location


Project Teams

London, UK:
Project Management and support
Hardware supply and installation
Data Centre and IPv6 connectivity

Nile University, Egypt:
Software Programming


Crawler (back
end)

Crawler Model
Name (eth0)
HP DL360p
turtle.ipv6matrix.org ;
crawler.ipv6matrix.org
turtle.ipv6matrix.com ;
crawler.ipv6matrix.com
turtle.ipv6matrix.net ;
crawler.ipv6matrix.net
IPv4 address 212.124.204.162 / 100 Mb/s
(eth0) /
speed
IPv6 address 2a00:19e8:20:1::a2 / 100 Mb/s
(eth0) /
speed
Name (eth1) shell.ipv6matrix.org
IPv4 address 194.33.63.250 / 1 Gb/s (GIH
(eth1) / private address space)
speed
CPU 2 x Dual Core Intel(R)
Xeon(TM) CPU 3.60GHz
RAM 4 Gb DDR2 SDRAM
HD Storage 146 Gb hardware SATA 2-disk
RAID (hot swappable)
PSU 2 x hot-swappable redundant
535W.
Operating CENTOS 5 Linux / updated
System

Web Server
(front end)

Web Server Model
Name (eth0)
HP DL140
elephant.ipv6matrix.org ;
www.ipv6matrix.org
elephant.ipv6matrix.com ;
www.ipv6matrix.com
elephant.ipv6matrix.net ;
www.ipv6matrix.net
IPv4 address 212.124.204.170 / 100 Mb/s
(eth0) /
speed
IPv6 address 2a00:19e8:20:1::aa / 100 Mb/s
(eth0) /
speed
Name (eth1) tusk.ipv6matrix.org
IPv4 address 194.33.63.251 / 1 Gb/s (GIH
(eth1) / private address space)
speed
CPU 2 x Dual Core Intel(R) Xeon(TM)
CPU 3.40GHz
RAM 4 Gb DDR2 SDRAM
HD Storage 2 x 1 Tb fast SATA
PSU Single 500W
Operating System Ubuntu 4.4 Linux / updated


Router
Router

Model CISCO 2811

Operating Advanced IP Services IOS
System
DRAM 64 Mb

Ethernet Ports / 2 / 100 Mb/s
speed
Interface card / MN-16ESW 16 port / 100
speed Mb/s


Local Network


Software flowchart


Web Site
Structure


www.ipv6matrix.org Web Site


What are we tracking?
Host IPv6 penetration
Using IPv4 Geo-location coordinates
Includes generic TLD (gTLDs, .com, .net, .org) and
country code ccTLDs
Two types of information:
Infrastructure: DNS + Web + E-mail + Time server
(NTP)
This tracks all of the infrastructure required to run
IPv6 Web services
Web sites only (actual content)
This tracks the Web services themselves. Usually this
percentage is lower than the figure for the
infrastructure


What are we archiving?
Everything that we are tracking on the previous page,
plus:
Testing of connectivity to the above services in case
IPv6 addresses are advertised but no service runs on
them
Tracing of route from London Docklands to each one
of these hosts both using IPv4 and IPv6 – and
archiving all of this information in text format
A lot more data, accessible from the Web site archive
In April 2011, the size of this database in text format is
approx 60Gb and continually increasing (the testing
software runs 24/7)


Data Archives

Selecting Data Archives provides
a link to the detailed information
Table.
Click on “Search”, to go
directly to the Top Level Domain,
or scroll down to the desired
Top Level Domain.


Data Archives

Example Data in Data Archives (stop
at date)

Click on the “+” to the left of the
Top Level Domain to gain
access the sub-menu containing
the dates of all the data runs,
and so on, to reveal results.

year-month-day__hour-min-sec


Data Archives - results

Click on the links to open a
new window containing
all of the detailed results
formatted in a table.


Data Archives - results
Basic IPv4/IPv6 connectivity table

Domain penetration summary table

Geographical IP database

Type of IPv6 access (tunnels etc.)

Trace path, hop count, MTU IPv4 & IPv6

Ping count IPv4 and IPv6

Reverse IPv4 and IPv6

SOA for Nameserver

Service detection for SMTP, HTTP and HTTPs

TLS detection for SMTP

Detailed Data for .AE
Clicked on tcp80_WWW_ae: connection to http port


Using the filter

Use the filter to select by Top Level Domain, or by type of service tested.


Results

November-December 2011


Europe Data snapshot
1.56% 5.44%
2.02% 5.68%
IPv6 Host Penetration
Infrastructure 6.86%
7.91%
- DNS 4.47% 2.02% 5.58%
4.75% 2.13% 5.39%
- Web 8.98%
2.86% 8.88% 5.48% 13.70% 0.88%
- E-mail 2.29%
2.80% 4.06% 9.34% 6.44% 6.72%
4.00%
- NTP 4.04%

3.53% 2.99%
2.71% 0.83%
Low Sample 14.74% 10.92%
Low accuracy 15.14% 11.83%
0.47% 0.60% 20.36%
9.97% 0.64% 0.58% 18.54%
July 2011 9.82%
16.43% 10.72%
6.31% 16.26% 11.09%
December 2011 6.14% 16.25%
16.66% 5.98% 2.69% 8.07% 2.41%
6.30% 2.67% 10.01% 2.33%
3.01%
4.34%

Europe Data snapshot
0.48% 2.50%
0.47% 2.60%
0.26%
Web only 0.51%
1.07% 0.94% 0.93%
0.99% 1.17% 0.94%
2.07%
0.64% 1.85% 1.68% 7.15% 2.19% 0.91%
0.71% 2.41% 1.68% 7.45% 0.86%
0.19%
0.19%

0.28% 2.47%
0.27% 0.95%
Low Sample 0.87% 0.41%
Low accuracy 1.00% 0.42%
0.32% 0.24% 26.58%
12.01% 0.32% 0.29% 26.70%
July 2011 11.67%
12.48% 0.34%
December 2011 8.16% 12.29% 0.48%
9.14% 4.08%
4.15% 0.77% 1.32% 3.71% 0.63%
0.85% 1.30% 4.00% 0.65%
1.71%
1.59%

Country DNS+WEB+E-mail+NTP % change
July 2011 Dec 2011

Moldova (*) 20.36% 18.54% -1.82%
Switzerland 16.25% 16.66% 0.41%
Slovakia 16.43% 16.26% -0.17%
France 14.74% 15.14% 0.40%
Finland 8.98% 13.70% 4.72%
Ukraine 10.02% 11.83% 1.81%
Slovenia 10.93% 11.33% 0.40%
Poland 10.72% 11.09% 0.37%
Czech Republic 8.07% 10.01% 1.94%
Portugal 9.97% 9.82% -0.15%
Netherlands 8.88% 9.34% 0.46%
Denmark 6.86% 7.91% 1.05%
Bosnia and Herzegovina (*) 7.66% 7.51% -0.15%
Latvia 0.88% 6.72% 5.84%
IPv6 Host Penetration Germany
Austria
5.48%
5.98%
6.44%
6.30%
0.96%
0.32%
Spain 6.31% 6.14% -0.17%
Croatia 5.68% 5.83% 0.15%
Norway 5.44% 5.68% 0.24%
Infrastructure Estonia 5.58% 5.39% -0.19%
Sweden 5.44% 4.75% -0.69%
- DNS or Greece 3.91% 4.34% 0.43%
Belgium 2.86% 4.06% 1.20%
- Web or Lithuania 4.00% 4.04% 0.04%
- E-mail or Macedonia 4.04% 3.33% -0.71%
Luxembourg 3.21% 3.21%
- NTP Serbia (*) 3.09% 3.06% -0.03%
Ireland 2.29% 2.80% 0.51%
United Kingdom 3.53% 2.71% -0.82%
Italy 2.69% 2.67% -0.02%
(*) Low Sample Hungary 2.41% 2.33% -0.08%
Low accuracy Russia 2.02% 2.13% 0.11%
Iceland 1.56% 2.02% 0.46%
Belarus 2.99% 0.83% -2.16%
Romania 0.60% 0.58% -0.02%
Bulgaria 0.71% 0.50% -0.21%


Country WEB % change
July 2011 Dec 2011

Moldova (*) 26.58% 23.70% -2.88%
Slovakia 12.48% 12.29% -0.19%
Portugal 12.01% 11.67% -0.34%
Spain 8.16% 9.14% 0.98%
Germany 7.15% 7.45% 0.30%
Slovenia 3.96% 4.77% 0.81%
Macedonia 5.43% 4.23% -1.20%
Switzerland 4.08% 4.15% 0.07%
Czech Republic 3.71% 4.00% 0.29%
Norway 2.50% 2.60% 0.10%
Belgium 1.85% 2.41% 0.56%
Finland 2.07% 2.19% 0.12%
Netherlands 1.68% 1.68% 0.00%
Croatia 1.55% 1.63% 0.08%
Greece 1.71% 1.59% -0.12%
IPv6 Host Penetration Italy 1.32% 1.30% -0.02%
Russia 0.94% 1.17% 0.23%
France 0.87% 1.00% 0.13%
Sweden 1.07% 0.99% -0.08%
Web only Belarus 2.47% 0.95% -1.52%
Estonia 0.93% 0.94% 0.01%
Latvia 0.91% 0.86% -0.05%
Austria 0.77% 0.85% 0.08%
Ireland 0.64% 0.71% 0.07%
Hungary 0.63% 0.65% 0.02%
(*) Low Sample Denmark 0.26% 0.51% 0.25%
Low accuracy Poland 0.34% 0.48% 0.14%
Iceland 0.48% 0.47% -0.01%
Ukraine 0.41% 0.42% 0.01%
Romania 0.24% 0.29% 0.05%
United Kingdom 0.28% 0.27% -0.01%
Lithuania 0.19% 0.19% 0.00%
Bulgaria 0.31% 0.16% -0.15%
Bosnia and Herzegovina (*) 0.00% 0.00% 0.00%
Serbia (*) 0.00% 0.00% 0.00%

Europe Trends
Many countries see a significant rise in
infrastructure
Likely to be caused by a main hosting
provider, installing dual-stack Name-servers
Slovakia & Portugal still leading with dual-
stack Web sites
Spain catching up.
Country with largest number of dual stack
Web sites in Europe: Germany
IPv6 Day in June 2011 might have
generated positive results


Asia Data snapshot
IPv6 Host Penetration 2.02%
2.13%
15.89%
Infrastructure 12.90%
- DNS or
- Web or
- E-mail or 0.18% 1.69%
0.48% 0.20% 1.83%
- NTP 0.50%
2.85% 0.30%
2.76% 0.22% 0.25%
0.54% 2.21%
1.87% 2.38% HK: 2.61%
Low Sample 0.00% 2.21% 3.58%
0.00% 3.52%
Low accuracy
0.65% 11.32%
1.92% 0.64% 11.94%
1.84% 3.03% 22.73%
July 2011 3.12% 5.90% 22.73%
6.16%
1.23% 3.42% 0.10%
1.22% 4.22% 0.10% 3.80% 3.65%
December 2011 3.47% 4.16%

SG: 25.52%
SG: 25.52%


Asia Data snapshot
1.17%
4.29%
3.49%
Web only

0.16% 0.23%
0.46% 0.15% 0.28%
0.38%
1.83% 0.30%
1.77% 0.14% 0.13%
0.14% 1.44%
0.48% 1.49% HK: 0.71%
Low Sample 0.00% 0.38% 4.46%
0.00% 4.69%
Low accuracy
0.69% 11.76%
3.70% 0.65% 11.21%
3.57% 3.03% 50.00%
July 2011 3.12% 1.03% 50.00%
1.14%
1.23% 0.88% 0.11%
1.22% 1.24% 0.11% 0.99% 0.95%
December 2011 0.91% 0.97%

SG: 0.00%
SG: 0.65%


July 2011 Dec 2011

Singapore 25.52% 31.32% 5.80%
Armenia (*) 15.89% 12.90% -2.99%
Sri Lanka (*) 11.32% 11.94% 0.62%
Indonesia 5.90% 6.16% 0.26%
Malaysia 3.42% 4.22% 0.80%
New Zealand 3.65% 4.16% 0.51%
Philippines (*) 3.58% 3.52% -0.06%
Australia 3.80% 3.47% -0.33%
Oman (*) 3.03% 3.12% 0.09%
IPv6 Host Penetration Saudi Arabia (*) 2.85% 2.76% -0.09%
Hong Kong 2.61%
Taiwan 2.21% 2.38% 0.17%
Thailand 1.87% 2.21% 0.34%
Infrastructure Russia 2.02% 2.13% 0.11%
- DNS or Qatar (*) 1.92% 1.84% -0.08%
- Web or Japan 1.69% 1.83% 0.14%
Pakistan 1.23% 1.22% -0.01%
- E-mail or United Arab Emirates (*) 0.65% 0.64% -0.01%
- NTP India 0.22% 0.54% 0.32%
Iran 0.48% 0.50% 0.02%
South Korea 0.30% 0.25% -0.05%
(*) Low Sample China 0.18% 0.20% 0.02%
Low accuracy Vietnam (*) 0.10% 0.10% 0.00%
Kuwait (*) 0.00% 0.00% 0.00%


July 2011 Dec 2011

Sri Lanka (*) 11.76% 11.21% -0.55%
Philippines (*) 4.46% 4.69% 0.23%
Qatar (*) 3.70% 3.57% -0.13%
Armenia (*) 4.29% 3.49% -0.80%
Oman (*) 3.03% 3.12% 0.09%
Saudi Arabia (*) 1.83% 1.77% -0.06%
Taiwan 1.44% 1.49% 0.05%
Malaysia 0.88% 1.24% 0.36%
Pakistan 1.23% 1.22% -0.01%
IPv6 Host Penetration Russia 0.94% 1.17% 0.23%
Indonesia 1.03% 1.14% 0.11%
New Zealand 0.95% 0.97% 0.02%
Australia 0.99% 0.91% -0.08%
Hong Kong 0.71% 0.71%
Web only
United Arab Emirates (*) 0.69% 0.65% -0.04%
Singapore 0.00% 0.65% 0.65%
Thailand 0.48% 0.38% -0.10%
Iran 0.46% 0.38% -0.08%
(*) Low Sample Japan 0.23% 0.28% 0.05%
Low accuracy China 0.16% 0.15% -0.01%
India 0.14% 0.14% 0.00%
South Korea 0.30% 0.13% -0.17%
Vietnam (*) 0.11% 0.11% 0.00%
Kuwait (*) 0.00% 0.00% 0.00%


Asia Trends

Singapore still leading by far in
infrastructure but trailing in dual-stack Web
Site content
China results abnormally very low – content
not dual stacked or IPv6 behind firewall?
India also low both in infrastructure and
most popular Web sites with dual stack
Less reliability of results due to smaller
sample size in many countries of the region


Africa Data snapshot
36.61%
Infrastructure
- DNS or
- Web or
- E-mail or
- NTP
16.67%
9.52% 6.64%
7.42%
Low Sample
10.20%
Low accuracy 10.20%

July 2011

December 2011
0.84%
0.96%


Africa Data snapshot
0.00%

Web only

10.00%
10.00% 2.63%
2.78%
Low Sample
14.29%
Low accuracy 14.29%

July 2011

December 2011
0.33%
0.41%


July 2011 Dec 2011

Tunisia (*) 36.24% 36.61% 0.37%
Tanzania (*) 10.20% 10.20% 0.00%
Senegal (*) 16.67% 9.52% -7.15%
Infrastructure Kenya (*) 6.64% 7.42% 0.78%
- DNS or South Africa 0.84% 0.96% 0.12%
- Web or
- E-mail or
- NTP The small number of hosts tested make the results
for most of Africa appear higher than they really are
(*) Low Sample
Low accuracy


July 2011 Dec 2011

Tanzania (*) 14.29% 14.29% 0.00%
IPv6 Host Penetration Senegal (*) 10.00% 10.00% 0.00%
Kenya (*) 2.63% 2.78% 0.15%
South Africa 0.33% 0.41% 0.08%
Web only Tunisia (*) 0.00% 0.00% 0.00%

(*) Low Sample The small number of hosts tested make the results
Low accuracy for most of Africa appear higher than they really are


Africa Trends

Very few countries with IPv6
connectivity
Dual Stack Islands starting to appear
Can be compared with the growth of
Internet connectivity in the nineties


Compare Historical data on
African Internet Connectivity

June 1994 May 1997
Source: Internetology - http://www.nsrc.org/codes/bymap/ntlgy/ntlgy.htm

North/South America
2.22%
2.42%

Infrastructure 4.79% 3.12%
5.81% 3.57%
- DNS or PR: 6.25%
PR: 7.14%
- Web or 0.56% 4.27%
- E-mail or 0.53% 4.05%
- NTP
100.00%
100.00%

0.93% 2.32%
Low accuracy
2.45%
3.70%

July 2011 0.30%
0.30%
0.00% 5.33%
December 2011 2.63% 0.48% 5.41%
0.49%
0.56%
0.57%


North/South America
0.34%
0.54%


Web only 0.26%
0.19%
PR: 6.25%
PR: 7.14%

0.97% 1.79%
0.86% 1.85%

0.00%
0.00%

0.29% 1.36%
Low accuracy

1.80%
July 2011 6.38%
0.31%
0.31% 0.49% 5.11%
December 2011 0.00% 0.24% 2.68%
2.44%
0.54%
0.56%


July 2011 Dec 2011

Guatemala (*) 100.00% 100.00% 0.00%
Puerto Rico (*) 6.25% 7.24% 0.99%
United States 4.79% 5.81% 1.02%
Uruguay (*) 5.33% 5.41% 0.08%
Venezuela (*) 4.27% 4.05% -0.22%
Ecuador (*) 2.45% 3.70% 1.25%
IPv6 Host Penetration Cuba (*) 3.12% 3.57% 0.45%
Peru (*) 0.00% 2.63% 2.63%
Canada 2.22% 2.42% 0.20%
Infrastructure Brasil 1.93% 2.32% 0.39%
- DNS or Panama (*) 0.84% 0.93% 0.09%
- Web or Chile 0.56% 0.57% 0.01%
- E-mail or Mexico 0.56% 0.53% -0.03%
- NTP Argentina 0.48% 0.49% 0.01%
Colombia (*) 0.30% 0.30% 0.00%
(*) Low Sample
Low accuracy For (*) the small number of hosts tested make the results
in many countries appear higher than they really are


July 2011 Dec 2011

Puerto Rico (*) 6.25% 7.14% 0.89%
Ecuador (*) 1.80% 6.38% 4.58%
Uruguay (*) 2.68% 2.68% 0.00%
Peru (*) 0.00% 2.44% 2.44%
Venezuela (*) 1.79% 1.85% 0.06%
Brasil 1.34% 1.36% 0.02%
IPv6 Host Penetration Mexico 0.97% 0.86% -0.11%
Chile 0.54% 0.56% 0.02%
Canada 0.34% 0.54% 0.20%
Web only Colombia (*) 0.31% 0.31% 0.00%
Panama (*) 0.23% 0.29% 0.06%
Argentina 0.28% 0.24% -0.04%
United States 0.26% 0.19% -0.07%
(*) Low Sample Guatemala (*) 0.00% 0.00% 0.00%
Low accuracy Cuba (*) 0.00% 0.00% 0.00%

For (*) the small number of hosts tested make the results
in many countries appear higher than they really are


America Trends
USA leads in infrastructure but low on dual
Stack Web site.
Results of IPv6 Day on infrastructure itself?
Brazil leading in Latin America
Low percentage of IPv6 Web Sites across the
region shows little impact of IPv6 day (June 8)
Elsewhere, data based on low number of
Web sites, needs to be taken in moderation
(the restricted number of hosting providers
can make figures jump several percentage
points)


Worldwide Trends
We are seeing a slow growth in dual stack IPv4-IPv6
implementation
The June 8th IPv6 Day appears to have had more
impact in several European Countries than elsewhere
A decrease in percentage in some countries, points to
unstable peering agreements (the IPv6 network is less
closely meshed than the IPv4 network)
Bearing in mind the Asia Pacific Regional Internet
Registry has run out of IPv4 addresses, it is alarming
to see so few Web sites up and running IPv6 in that
region, especially in countries where Internet growth is
high, such as in India.


Problems / Possible Errors

Lots of errors in the DNS – commas, no A, no AAAA
record, looping MX, etc.
Firewalls and security software:
Blocking of network segments
detecting denial of service attack (DoS) by error:
• Unusual UDP traffic. Trace-path / ping, SMTP, HTTP,
Secure HTTP, NTP port testing.
Internet snapshot from one location only
Less accurate results with small input data size (small
number of domains tested)
Disputed accuracy of Geographical IP database


Example: fake IPv6 AAAA record

Country specific results? /P2
These “ipv6” addresses are AAAA records
which pretend to denote an IPv6 address,
(examples from the report / finding
but are actually IPv4 addresses.
errors etc.)
There are plenty of examples of such mis-
configuration in the DNS.
Worse problems are caused by commas in domain
names, ie. www.domain,com as well as IP
addresses returning localhost 127.0.0.1 or ::1 etc.

Comma!

Perform a search for ipv6 field starting with ::ffff


Machine readable output

Command directly to SQL database

http://www.ipv6matrix.org:4444/getTable?crawl=2010-07-18_12-24-48&tld=com&table=WWW_com

It is therefore possible to use an entirely new user interface with the back end database


Future work / funding required

Add more domains to be tested
Add more features to be tested
Current front end Web Pages are only an
example of possible analysis
Develop new data visualisation
Perform further analysis
Perform historical/time analysis from archives
Develop an engine to write automated reports
Duplicate Crawler to other regions


World IPv6 Launch

Major Internet service providers
(ISPs), home networking
equipment manufacturers, and
web companies around the world
are coming together to
permanently enable IPv6 for their
products and services.
Target: 6 June 2012
http://www.worldipv6launch.org


Thank you
Contact: ISOC England – contact@isoc-e.org

CTM
International


IPv6 Matrix Exec Summary Dec 2011 Results - ICCA Pondicherry 31 Jan 2012

Recomendados

Recomendados

Mais conteúdo relacionado

Destaque

Destaque (8)

Semelhante a IPv6 Matrix Exec Summary Dec 2011 Results - ICCA Pondicherry 31 Jan 2012

Semelhante a IPv6 Matrix Exec Summary Dec 2011 Results - ICCA Pondicherry 31 Jan 2012 (20)

Mais de Olivier MJ Crépin-Leblond

Mais de Olivier MJ Crépin-Leblond (13)

Último

Último (20)

IPv6 Matrix Exec Summary Dec 2011 Results - ICCA Pondicherry 31 Jan 2012