0

Anycast secondary IPv6 routing to anexia-ath instead of zepto-lon

Evening,

Once again, IPv6 routing has gone awry from AANET, and this time it's routing the secondary anycast to anexia-ath instead of zepto-lon, which means it's 60 ms latency instead of 10 ms.

Why does this consistently keep happening to IPv6 which ends up with degraded latency as a result? I've never seen it happen with IPv4.

Is any fix in the pipeline for this odd behaviour? Presuming it's some kind of fallback, why does it end up routing to somewhere a long way away with high latency instead of somewhere nearer with better latency?

Diag: https://nextdns.io/diag/ba88acc0-715a-11f0-8da8-dd060582d10a

Cheers,

-Will

14 replies

null
    • Will_Tisdale
    • 3 wk ago
    • Reported - view

    Still the same today...need a fix please!

    • skeggox
    • 3 wk ago
    • Reported - view

    If you aren't running the nextdns-cli already it might be worth a try if you can.

    It resolved my issue (I'm in Canberra using AussieBB).

    https://help.nextdns.io/t/q6yk989/dns1-and-dns2-nextdns-io-not-resolving-and-incorrect-dns-server-location-selected

      • Will_Tisdale
      • 3 wk ago
      • Reported - view

       the nextdns cli doesn’t use anycast so isn’t affected by this issue.

      Your anycast routing is equally as broken looking at your report and that needs fixing too. The routing logic just doesn’t make sense a lot of the time.

      • skeggox
      • 3 wk ago
      • Reported - view

       Yeah I agree. I didn't want to wait for routing to be fixed for anycast.

    • Zach_B
    • 3 wk ago
    • Reported - view

    Could be worse.. lol.. I have the same issue.. Except anexia-ath is 149ms for me... 

    https://nextdns.io/diag/fc426b40-7265-11f0-b423-eb7c0ef70556

    • Will_Tisdale
    • 3 wk ago
    • Reported - view

    I just noticed on my monitoring that this is now resolved as of a few minutes ago and secondary anycast IPv6 is routing to anexia-lon as it should be. 

      • Will_Tisdale
      • 3 wk ago
      • Reported - view

       and then about 10 minutes later it went back to routing to anemia-ath. Why is this so hard to get fixed?

    • Will_Tisdale
    • 2 wk ago
    • Reported - view

    Still broken. 

    • Will_Tisdale
    • 2 wk ago
    • Reported - view

    The same issue still exists and it's been nearly a week since I reported it. Can it please be fixed or an explanation provided as to why it's taking so long to fix? Thanks!

    Current diag: https://nextdns.io/diag/1213daf0-76d8-11f0-b677-99e657676d12

      • Will_Tisdale
      • 12 days ago
      • Reported - view

       still broken, and my post with the updated diag from 4 days ago is still pending review.

      This is a paid service and the level of support for a simple issue that is previously been a quick fix is currently unsatisfactory. All I want is an explanation or a fix, nothing difficult.

    • Will_Tisdale
    • 12 days ago
    • Reported - view

    This is still broken and my post with the new diag from 4 days ago is still pending review.

    This is a paid service and the level of support for something that has previously been a quick fix is currently unsatisfactory.

    All I want is an explanation or a fix, nothing hard.

    Thanks.

    • Mane
    • 12 days ago
    • Reported - view

    Force the server, it is not recommended. Change the DoH to “https://anexia-lon-1.edge.nextdns.io/profileid

      • Will_Tisdale
      • 12 days ago
      • Reported - view

       I'm not using DoH. This is anycast IPv6 using the IP address that is broken. There's no way of "forcing" anything in that case.

    • Will_Tisdale
    • 5 days ago
    • Reported - view

    Still not fixed.

Content aside

  • 5 days agoLast active
  • 14Replies
  • 132Views
  • 4 Following