On Thu, Apr 16, 2026 at 08:56:16PM +0200, Marco Moock via tor-relays wrote:
Hello!
I am referring to https://metrics.torproject.org/rs.html#details/BBBBBAD453263D786EC34AB68A062...
I cannot reach this relay via the IPv6 address, a tcp traceroute to port 443 ends at 2a00:6020:ffff:ffff::14.
Can someone explain why this is listed as reachable in consensus?
I can reach this relay's IPv6 address from MIT: $ telnet 2a00:6020:b2c5:d8a9:bbbb:bad4:5326:3d78 443 Trying 2a00:6020:b2c5:d8a9:bbbb:bad4:5326:3d78... Connected to 2a00:6020:b2c5:d8a9:bbbb:bad4:5326:3d78. Escape character is '^]'. My home internet connection still doesn't have IPv6 support so I can't check from there. When I use the bridge scan reachability test (it's technically designed with bridges in mind, but it just does a TCP reachability test), it seems to work from there too: https://bridges.torproject.org/scan/scan?address=%5B2a00%3A6020%3Ab2c5%3Ad8a... And on the form at the bottom of https://consensus-health.torproject.org/#relayinfo I see that all the dir auths besides longclaw can reach it -- and longclaw can't reach gabelmoo either so I think maybe longclaw isn't doing IPv6 reachability testing at all. So, sounds like the problem might be on your end? --Roger