You are on page 1of 42

IPv6 Deployment Survey

Based on responses from the RIPE


community during June 2009
Maarten Botterman
RIPE 59, Lisbon, 6 October 2009
Why IPv6 Deployment Monitoring?
• The Internet has become a fundamental infrastructure,
worldwide, for economic and social activity, and its
usage continues to grow exponentially:
• More users
• New applications (eg mobile, RFID etc)

• The transition from IPv4 to IPv6 is the only sustainable


option, in the long run.
• A smooth transition requires understanding the
challenges, and a timely start.

2
IPv6 deployment monitoring
European IPv6 Action Plan
May 2008

ADVANCING THE INTERNET: Action Plan for the


deployment of Internet Protocol version 6 (IPv6) in
Europe
• Preparing for the growth in Internet usage and for
future innovation
• Maintaining Europe's competitiveness

• So … what can be done?


http://www.ipv6.eu/admin/bildbank/uploads/Documents/Commision/COM_.pdf
3
IPv6 deployment monitoring
Basically, it’s simple
• The European Commission and Europe’s Member
States are committed to support a smooth transition
towards IPv6, for clear public interest reasons
• … and think a significant step should be done by 2010
• Public sector procurement
• Monitoring security and privacy implications

• Yet the IPv6 transition will be driven by the Internet


community
• Providers
• Users
4
IPv6 deployment monitoring
Therefore this proposition was
made to the RIPE community
• How about making sure the European
Commission knows what could be done,
usefully, to help ensure that smooth
transition?

• Really understand the scope of the problem


• Identify the bottlenecks
• Propose useful steps to support the transition

5
IPv6 deployment monitoring
IPv6 Deployment Monitoring project:
putting the facts on the table
• Measuring:
– deployment in EU countries (% end users)
These are source address based on passive measurements
– availability (% IPv6 web-based services)
– differences between IPv4 and IPv6 performance
These are measurements on quality of service

• Information gathering:
– Global sources
– Key informant interviews
– IPv6 Survey

6
IPv6 deployment monitoring
What about the survey ?
• Aim is to establish the best possible comprehensive view of present
IPv6 penetration and future plans of IPv6 deployment
• Best way to establish this is to ask the Internet providers and users,
basically: the RIPE participants
• ARIN carried out such a survey with its members in March 2008, a
starting point for the currently proposed survey
– Survey was prepared and carried out by TNO/GNKS in close collaboration
with RIPE NCC
– Survey was kept short, and focused on essentials
– Privacy is guaranteed
• APNIC carried out the same survey during September 2009
– courtesy of APNIC we are able to compare some of the results, already in this
presentation!

7
IPv6 deployment monitoring
Response to questionnaire

31% completed
partial
57%
No attempt
12%
n=2000

610 repondents from 54 countries

source: TNO/GNKS 2009

8
Geographic spread responses
90
80
70
60
50
40
30
20
10
0

n = 610

source: TNO/GNKS 2009

9
Respondent categories

Education

Government

ICT/Internet
tools
Internet
content
ISP
ISP
Non ICT
62% industry
Other

R&D
n = 610

source: TNO/GNKS 2009


10
IPv6 presence respondents

No 37%

Yes, within internal networks 31%

Yes, on the Internet 49%

n = 464

0% 10% 20% 30% 40% 50% 60%

source: TNO/GNKS 2009


11
IPv6 presence respondents
APNIC (295)
RIPE (464) 33%
No
37%

43%
Yes, within internal networks
31%

50%
Yes, on the Internet
49%

0% 10% 20% 30% 40% 50% 60%

source: TNO/GNKS 2009


11
More profiling
• ~ 90% of respondents have a registration
services agreement with RIPE NCC
• ~80% if respondents works for profit
• ~75% is EU based
• ~85% of 356 ISPs has less than 100,000
customers
• ~50% of respondents have less than 50
employees
12
IPv6 vs IPv4 traffic

16%

IPv6 traffic is insignificant

IPv6 traffic is less than IPv4 traffic

IPv6 traffic is same as IPv4 traffic

IPv6 traffic is greater than IPv4 traffic

82% n = 610

source: TNO/GNKS 2009


13
IPv6 vs IPv4 traffic
RIPE APNIC

16%
17%

82% 77%

IPv6 traffic is insignificant

IPv6 traffic is less than IPv4 traffic

IPv6 traffic is same as IPv4 traffic


source: TNO/GNKS 2009 13
IPv6 traffic is greater than IPv4 traffic
EU sector consider having IPv6
allocation
100%
90%
80%
70%
60%
50%
40%
30%
20%
No
10%
Yes
0%

n = 456

source: TNO/GNKS 2009


14
EU ISPs consider having IPv6
allocation
180
160
140
120
100
80
No
60 125
Yes
40
49
20
25 20
0
'Small (up to 50 Medium (51 up to 'Large (251 up to 'Very Large (more
employees) 250 employees) 2,500 employees) than 2,500
employees)

source: TNO/GNKS 2009


15
ISP, do you considering promoting IPv6
uptake to your customers
100%

90%

80% 43%
70%

60% No
50% 14% Maybe
40% Yes

30% n = 380

20% 43%

10%

0%

source: TNO/GNKS 2009


16
ISP, do you considering promoting IPv6
uptake to your customers
100%
13%
90%

80% 43%
70% 34%
60%

50%
No
14%
Maybe
40%
Yes
30%
54%
20% 43%
10%

0%

RIPE APNIC

source: TNO/GNKS 2009


16
Why not considering IPv6?
80%
70%
60%
50%
40%
30%
20%
10%
0%

n = 467

source: TNO/GNKS 2009

17
Why not considering IPv6?
80%
70%
60%
50%
40%
30%
20%
10%
0%

RIPE APNIC

source: TNO/GNKS 2009


17
Biggest hurdles
Other
Business case
Information security those who
Availability of knowledge implement IPv6
Vendor support (467)
Costs

Other
those who don't
Business case
plan IPv6, yet
Information security (143)
Availability of knowledge
Vendor support
Costs

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

18
source: TNO/GNKS 2009
Biggest hurdles
Other
Business case
Information security those who
implement IPv6
Availability of knowledge
Vendor support
Costs ARIN
APNIC

RIPE
Other
Business case
Information security those who don't
plan IPv6,yet
Availability of knowledge
Vendor support
Costs
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

18
source: TNO/GNKS 2009
Main drivers to IPv6 deployment
80%

70%

60%

50%

40%

30%

20%

10%

0%
Customer Availability of Want to Want to be Make sure Other
demand IPv4 address benefit from "ahead of the IPv6 is
space advantages game" supported in n = 435
asap our products

source: TNO/GNKS 2009 19


Main drivers to IPv6 deployment
80%

70%
RIPE APNIC
60%

50%

40%

30%

20%

10%

0%
Customer Availability of Want to Want to be Make sure Other
demand IPv4 address benefit from "ahead of the IPv6 is
space advantages game" supported in
asap our products

source: TNO/GNKS 2009 19


Planning IPv6 deployment
100% No plan
90%
80% > 4 years
70%
60% > 2 years
50%
40% > 1 year
30%
'0,5 to 1 year
20%
10%
1 to 6 months
0%
Currently
deployed
n=610

source: TNO/GNKS 2009


20
100%
Planning IPv6 deployment
90%
80%
70%
60%
50%
RIPE
40%
30%
20%
10%
0%
Internal IPv6 transit IPv6 Peering DNS services Desktops Webservices Hosted IPv6 E-mail Cable/DSL
network

100%
90%
80%
70%
60%
50%
APNIC
40%
30%
20%
10%
0%

source: TNO/GNKS 2009 20


Experience: biggest problems with
IPv6 in production?
Lack of user demand

No experience, yet

Technical problems

Budget issues

Other

0% 10% 20% 30% 40% 50% 60%

source: TNO/GNKS 2009 21


Experience: biggest problems with
IPv6 in production?
Lack of user demand

No experience, yet

Technical problems

Budget issues APNIC


RIPE
Other

0% 10% 20% 30% 40% 50% 60%

source: TNO/GNKS 2009 21


Experience: Set-up
• Overwhelmingly dual-stack (~90%)
• Mostly native IPv6
3% 2%
4%

Native IPv6
12%
Tunnel (not automatic)

Automatic tunneling

Address translation
78% Other

22
Experience: Set-up
• Overwhelmingly dual-stack (~90%)
• Mostly native IPv6

3% 2% 8%
4%

9% Native IPv6
12% 3%
Tunnel (not automatic)

Automatic tunneling
50%

30% Address translation


78%
Other

RIPE APNIC
22
Indication of real IPv6 usage?

source: TNO/GNKS 2009


Ratio of IPv6/total IP visitors from selected countries
23
Indication of real IPv6 usage?

Ratio of IPv6/total IP visitors from EU countries to a measured web site

source: TNO/GNKS 2009

24
Growth foresight ?
200

180

160

140

120

100

80

60

40

20

0
AT BE BG CY CZ DE DK EE ES FI FR GR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK

Number of advertised IPv4 addresses per 100 inhabitants


source: NII/TNO/GNKS 2009 25
IPv4 address shortage indication?
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
AT BE BG CY CZ DE DK EE ES FI FR GR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK US

Ratio of announced/allocated of IPv4 addresses in the EU and US


(nb: announced is not necessarily assigned)
source: NII/TNO/GNKS 2009
26
Main conclusions, overall
• Much more IP addresses will be needed during the
coming years
• If only for other countries to get up to the IP maturity levels of
Scandinavian countries
• But also mobile internet, and Internet of Things
• Whatever happens: no new IPv4 addresses available
anymore, anywhere, at some point!
• IANA source IPv4 will be depleted by 2011
• RIPE source of IPv4 will be depleted by 2012/2013
• In some countries the need for new IP addresses will be greater
than in others

27
http://www.ipv6monitoring.eu/
Main conclusions survey 2009 (1/2)
• Need to be careful with drawing conclusions
• In particular when breaking down the sample to small size categories
• This group is biased by IPv6 interest, at least
• How well do respondents know what is really going on within their
organisation
• Overall reason for IPv6 not being a priority yet is “lack
of business case/lack of customer demand”
• IPv6 vendor support is still lacking
• How to turn this around?
• 63% of RIPE respondents have, or consider having an
IPv6 allocation, today
• Only 53% of Government respondents consider having IPv6
• Still 21% of all respondents not convinced of the need
to have IPv6 towards the future: why?
28
http://www.ipv6monitoring.eu/
Main conclusions survey 2009 (2/2)
• ISPs:
• 82% has, or considers having IPv6
• 56% has IPv6 in production
• 37% of ISPs in Europe do not consider IPv6 promotion
• What is needed to get more ISPs on board

• Web site content


• Just one out of 27 x Top 30 websites measured supports IPv6
• Set-up today is overwhelmingly dual stack and native
IPv6
• Are these the “lucky few”? Much less extreme “dual stack” in APNIC
region
• What future for tunneling, address translation?

29
http://www.ipv6monitoring.eu/
We thank all respondents for
their contributions !
• More than 70% indicated their willingness to
collaborate to further follow up questions
• More than 90% indicated their willingness to
respond again, next year

30
http://www.ipv6monitoring.eu/
This survey could not have been done
without the help of RIPE NCC, and APNIC
Thanks to the European Commission who has made this possible by
granting GNKS Consult and TNO a study contract on IPv6
Deployment, in line with the EU IPv6 Action Plan
Thanks to all RIPE members that helped improve the survey
instrument, before it was launched.
Thanks to RIPE and APNIC staff for support and help, and for
sending out the survey to their mailing lists.
Special thanks to KC Claffy (CAIDA), Karine Perset (OECD), Leslie
Daigle (ISOC), Paul Rendek and Nick Hyrka (RIPE NCC), Miwa Fujii
and Paul Wilson (APNIC) for their feedback, advice and support.

http://www.ipv6monitoring.eu/
http://www.ipv6monitoring.eu/

Questions regarding the survey and


this presentation:
Maarten Botterman
maarten@gnksconsult.com
The European IPv6 Web Site

http://ec.europa.eu/information_society/policy/ipv6

Questions regarding the Action Plan to the


European Commission:
Jacques.babot@ec.europa.eu
••• 45