IPv6 Subnet-Router Anycast (SRA) Dataset

This website provides the IPv6 SRA dataset, a collection of IPv6 router addresses discovered using Subnet-Router Anycast (SRA) probing. In addition to a weekly updated dataset, we offer a dashboard that compares our dataset with other well-known IPv6 address datasets such as the TUM Hitlist. We also provide information about routing loops on the IPv6 Internet to help avoid collateral damage during scanning.

141M

Observed IPv6 router addresses using subnet-router anycast probing

16k

Autonomous systems connect routers from the SRA datasets

Last scan: 2026-06-10

87.21% 123M

Additional IPv6 router addresses compared to other data sources

64.38% 91M

IPv6 router addresses replying to a direct ICMP Echo request (ping)

Column schema for .csv.zst
# IPv6 router addresses, scanned on YYYY-MM-DD
router_address  string
ping_success    bool
Column schema for .parquet
router_address  string
ping_success    bool
scan_date       date
                    

Comparison with other IPv6 address sources

TU Munich Hitlist

Latest dataset from: 2026-06-06 SRA dataset from: 2026-06-10

IPv6 Addresses


Total

33M

Overlap with SRA

12.71% 18M

Autonomous Systems


Total

22k

Overlap with SRA

93.09% 15k

CAIDA ITDK

Latest dataset from: 2026-06-11 SRA dataset from: 2026-06-10

IPv6 Addresses


Total

38M

Overlap with SRA

0.12% 170k

Autonomous Systems


Total

36k

Overlap with SRA

99.52% 16k

RIPE Atlas

Latest dataset from: 2026-06-06 SRA dataset from: 2026-06-10

IPv6 Addresses


Total

239k

Overlap with SRA

0.02% 35k

Autonomous Systems


Total

16k

Overlap with SRA

67.71% 11k

We analyze known issues in IPv6: Routing loops and amplification.

If you use our data, please use the following reference:

@article{khnsw-siius-25,
  author = {Maynard Koch and Raphael Hiesgen and Marcin Nawrocki and Thomas C. Schmidt and Matthias W{\"a}hlisch},
  title = {{Scanning the IPv6 Internet Using Subnet-Router Anycast Probing}},
  journal = {Proceedings of the ACM on Networking (PACMNET)},
  pages = {50:1-50:15},
  volume = {3},
  number = {CoNEXT4},
  year = {2025},
  month = {December},
  publisher = {ACM},
  address = {New York},
  url   = {https://doi.org/10.1145/3768997},
}