0

Improper Routing Since Weekend

Hello,

For several years the preferred routing for Anycast and ULL in Las Vegas, NV using Cox Communications has been with vultr-lax, zepto-lax, and anexia-lax. This weekend anexia-lax was dropped and replaced by both frantech-las and smarthost-las, which are both much slower than anexia-lax. Anexia-lax is still shown, but not being assigned anymore.

Here is the Diag:

https://nextdns.io/diag/d0393870-6422-11ed-a7df-2fa8c464d840

Please return to the previous arrangement.

Thank you.

6 replies

null
    • NextDNs
    • 2 yrs ago
    • Reported - view

    The diag shows you are going on LAX. What is the issue?

    • S_S.2
    • 2 yrs ago
    • Reported - view

    Yes, that moment in time I was routed to an lax server. The issue is that zepto-lax, vultr-lax, and anexia-lax are consistently good in the 10-20 ms range. However, smarthost-las often spikes to 40-60 ms and frantech-las is erratic and often spikes to 70-95 ms. Using DOH I may end up on either Anycast or ULL. By dropping anexia-lax and using the two las servers, I may be routed to those and at times when they are much slower. This was all avoided by using anexia-lax and avoiding the two las servers for all four of the server assignments.

    Thank you.

      • NextDNs
      • 2 yrs ago
      • Reported - view

      S S you seem to be setup with anycast, so there no chance that you will be steered to las, as they are not anycast pops.

    • S_S.2
    • 2 yrs ago
    • Reported - view

    I have a number of computers and devices that connect in different ways to NextDNS. Attached is a current Ping test from Windows showing a connection via ULL on smarthost-las with frantech significantly slower at 39 ms vs anexia-lax at only 13 ms. As I mentioned frantech-las often spikes to the 70-95 ms range. My tablets/iDevices often connect via ULL as well as Anycast.

    Here is another Diag using smarthost-las:

    https://nextdns.io/diag/8dec4ee0-667e-11ed-a7df-2fa8c464d840

    In summary, I am steered to Anycast and ULL, so the suboptimal ULL routing definitely matters. That is why this way changed by your Tech Support for this area shortly after you came out of Beta. Why it recently reverted is the question.

    Please correct this.

    Thank you.

    • S_S.2
    • 2 yrs ago
    • Reported - view

    Addendum:

    Friday morning the smarthost-las ping is now 385 ms in the same city. Anexia-lax is a very respectable 17 ms.

    Diag: https://nextdns.io/diag/6e26f820-6747-11ed-83c9-779a10e97d88

    Please take a look. This has been going on since last weekend as noted above.

    Thank you.

    • H_e
    • 2 yrs ago
    • Reported - view

    I am also having recurrent problems in the same city as the OP with slow resolution times that were not an issue before. It occurs when smarthost-las is being used. Look at the ULL Secondary in the following Diag where there is a huge jump between the final two hops and it ends up at over 200 msec.

    https://nextdns.io/diag/1753ae70-6900-11ed-ad55-4b032d3d1e68 

    This started last week.

    Please fix this!

Login to reply

Content aside

  • 2 yrs agoLast active
  • 6Replies
  • 159Views
  • 3 Following