Transcript IPv6 Readiness Measurement BoF Report

IPv6 Readiness Measurement
BoF Report
Ai-Chin Lu, TWNIC
September 19, 2014@AMM, APNIC 38
1
Major events/work items of BoF
2013/8
1st
BoF meeting
(APNIC, .tw, .cn, .jp, .nz)
Discuss methodologies
of readiness
measurement
2013/9~
2014/1
Global ipv6 Summit in
Taiwan –Ipv6
measurement workshop
(APNIC,
IRINN, .jp, .th, .tw, KISA)
Set up mailing
List &Build up website
2014/2
2nd BoF meeting
(APNIC, RIPE
NCC, .tw,.vn, NTT, BUPT,
Akamai, Cisco)
Propose 3 common
criteria
Classify different
measurement criteria
2
3 Common measurement criteria
• IPv6 Allocation and BGP advertisement
– Propose to use the percentage of ASes that
announce IPv6 Prefix and source data from
RIPENCC
• User Availability
– Propose to use the Google and APNIC data
• Service Availability
– Propose to use the DNS record of each ccTLD to
check the IPv6 availability of WEB/DNS/MAIL.
3
Major events/work items of BoF
2014/2~
2014/8
APAN 38 –Ipv6
measurement
workshop
(APNIC, .jp, .hk, .tw)
Post processing
Develp an ipv6-enable
Server detection tool
2014/9
3rd BoF meeting
(Cisco, APNIC,NTT,
sg,.tw, CHTL)
Propose different
methodology for
service availability
Participating
organizations
APNIC, RIPENCC,
Cisco, CHTL, NTT,
Akamai, BUPT,
CUHK,
JP, CN, VN, TW,
KISA, IRIN, SG, NZ,
TH
4
3 Common measurement criteria
1. IPv6 Allocation and BGP advertisement (# of ASN with IPv6 advertisement/# of ASN)
R
1
2
3
4
5
6
7
8
9
10
11
12
13
ccTLD
JP
SG
MY
TW
NZ
HK
AU
TH
ID
IN
CN
VN
KR
%
40.94%
39.15%
35.61%
31.54%
31.03%
24.70%
20.62%
20.24%
17.68%
17.31%
15.47%
12.93%
3.60%
2.User Availability
R
1
2
3
4
5
6
7
8
9
10
11
12
13
ccTLD
SG
JP
MY
AU
NZ
CN
TW
TH
HK
IN
ID
KR
VN
%
5.714%
5.418%
3.673%
1.291%
1.087%
0.624%
0.443%
0.220%
0.176%
0.070%
0.059%
0.009%
0.002%
5
3 Common measurement criteria
3.Service Availability
Propose to use the DNS record of each ccTLD to check the IPv6 availability of
WEB/DNS/MAIL.
Country
# of ccTLD
domain name
# of IPv6
enable DNS
The ratio of
# of IPv6
IPv6 enabled enabled
DNS
websites
The ratio of
The ratio of
# of IPv6
IPv6 enabled
IPv6 enable
enabled email
websites
email
TAIWAN
VIET NAM
531,127
278,364
2623
622
0.494%
0.223%
9687
3072
1.82%
1.10%
296
622
0.06%
0.22%
SINGAPORE
158,703
5851
3.687%
2571
1.62%
4458
2.81%
NEW ZEALAND
552,411
79160
14.330%
6242
1.13%
35299
6.39%
The issues of data collection
1. only 4 ccTLD registries provide the data
2. Some registries think it is not the right time to publish the data.
6
The 3rd IPv6 Readiness Measurement BoF
• 7 reports in IPv6 Readiness Measurement BoF:
– Ai-Chin Lu (TWNIC), BoF Status Report
– George Michaelson (APNIC), APNIC IPv6 measurement
– Mon Perez (SGNIC), Statistics of IPv6 Adoption in .SG
Names
– Tomohiro Fujisaki(NTT), Japan IPv6 Measurement
– Ching-Heng Ku(TWNIC), Taiwan IPv6 Measurement
– Eric Vyncke (CISCO), Cisco IPv6 Measurement
– Li-Fan Wu (CHT Laboratories), Chunghwa Telecom IPv6
Readiness Update
7
Current Status Report by
SGNIC、NTT、CHTL
• the statistics of IPv6 adoption in .SG Names ; the observation,
registrants rely on hosting providers to enable IPv6 for their
services, and service providers play a big role in the proliferation
of IPv6 websites.
• NTT reported the summary of IPv6 deployment status in Japan,
such as Core Network , Applications, Access Network, Users ,
Products, and Government IPv6 service.
• In Japan, IPv6 Internet connectivity has deployed smoothly, but
need to tackle Content services and Cellular services.
• CHT Laboratories reported CHT’s challenges, the IPv6 milestone,
and long term planning based on the IPv6 readiness
measurement report.
8
User Availability by APNIC、Cisco
APNIC
Cisco
• George Michaelson APNIC reported • Predict the trend of IPv6 user availability
that APNIC had longterm experiment
measuring end-user capability and
also decided to adopt google chart API
with a simple ‘timeline’ interface.
• http://stats.labs.apnic.net/ipv6/
9
Service Availability by Cisco、TWNIC
TWNIC
Cisco
•WEB service availability by Cisco Lab
•Cisco measures the top-100 sites accessed
from a country
• Web sites (from Alexa list) reachable over
IPv6
• http://6lab.cisco.com
•data collection is from Alexa’s 1 million
websites at http://www.ap-v6readiness.org
WEB service availability
DNS service availability
Mail service availability
10
Different methodologies of service availability
Sampling
size
M1
The # of ccTLD
Domain Name
TW Those under ccTLD
from the Alexa’s Top 1
million websites
Measurement
target
scope
Belonging to a
country based
organization
WEB
DNS
MAIL
Belonging to a
country based
organization
WEB
aggregated domain names
Domain names with the same 3rd level suffix should
not be counted twice if we want to calculate the
approximate numbers of domain name servers
through the Alexa’s approach.
For example
www.twnic.net.tw
rs.twnic.net.tw
CISCO
Alexa’s Top 100
websites
the top-100 sites
accessed from a
country
Methodology
Sevice Availabili ty of ccTLD 
WEB Availabili ty 
# of IPv6 Enabled Web/DNS/ Mail Servers
# of Domain Name records in ccTLD
# of IPv6 Enabled ccTLD Web Servers
# of ccTLD Web Servers in Alexa
DNS
MAIL
DNS Availabili ty 
# of IPv6 Enabled ccTLD DNS Servers
# of Aggregated ccTLD DN in Alexa
mail Availabili ty 
# of IPv6 Enabled ccTLD Email Servers
# of Aggregated ccTLD DN in Alexa
WEB
WEB Availabili ty 
# of IPv6 Enabled Web Servers
# of Web Servers in Alexa
11
Conclusion
• For service availability, We need to continuously
collect the data of different methodologies, and
watch closely.
• We encourage ccTLD registries to continuously
provide their IPv6 service availability data.
• we still welcome different methodology and
criteria of readiness measurement are proposed.
• We hope to cooperate with more organizations
to find out more appropriate measurement
criteria.
12
Thank you
Please join us:
Mailing list: [email protected]
Website: http://www.ap-v6readiness.org
13