There are several ways, how to measure IPv6 penetration – end user perspective (measured by Geoff Huston), infrastucture perspective (stats from BGP) or content provider perspective. The IPv6 penetration among content providers is usually measured by checking the appropriate resource records (RR) in DNS.
Graphs on this site are created from our domain-database which contains approximately 12 mil. of domains. If there is a RR pointing to a www.<domain> – domain is considered as a web domain and A and AAAA records for the domain are checked once a day. Alternative records for IPv6 (e.g. {www6, ipv6, www.v6) are checked as well.
Percentage of sites supporting IPv6
AAAA record announced for www.<domain>
Actual ratio of IPv4 only, dualstack and IPv6 only sites
Percentage of sites supporting IPv6 by TLD
Only TLD with more then 10 000 records are displayed.
Quality of IPv6 connectivity for site announcing AAAA record
Web domains supporting both protocols are checked for quality of connection. Using IPv4 and IPv6, we try to connect to the remote web server. The time between the first packet initiating relation (SYN) and the answer from the server (SYN, ACK) is measured and stored in the database.
Delta round trip time (drrt) is computed as rtt6 (round trip time over IPv6) – rtt4 (round trip time over IPv4). Results are clasiffied into 5 categories:
- Excellent: IPv6 works better then IPv4. drrt is < -50 ms.
- Same: The quality of IPv6 connection is the same as IPv4. drrt is in <-50 ms, 50 ms> interval.
- Poor: IPv6 has worse response time then IPv4. drrt is in <100ms, 2 s> interval.
- Very poor: IPv6 is very slow comparing to IPv4. drrt is > 2s.
- Invalid: There is no response over IPv6 even though the website annouces AAAA record.
Sites with broken IPv6 connectivity by TLD
Sites that announces AAAA record but we are not able to connect to the site using IPv6. This test is measured from our ASN. (AS 197451).