Has anyone seen a drop in magic DNS performance in the last month?
I’m having this in situations where the DNS would be getting hammered - for example my apt updates ansible script, or Uptime Kuma checks. If I switch to IP addresses it works fine.
You must log in or register to comment.
I went and edited my hosts file and added all of my devices, but I only have a handful. Tailscale on macOS has a lot of bugs, this being one of many.
Hate that you’re downvoted for this. Tailscale is incredible software but it is buggy as hell.
I gave up on 90% of the fancy features and just do most my routing from one node with good ol’
nftablesandip route/ip -6 route.
You’ve got a serious misconfiguration and are being rate-limited by their resolvers.
MagicDNS should only be resolving for devices on the Tailnet, not things like apt or Kuma. See here.
A Restricted Nameserver should be resolving for only your Tailscale Devices, and Global Server for everything else. You should never have normal services on your machines reaching out on Tailscale to query DNS unless you want the behavior you’re current seeing, which is unnecessary.
Where exactly do you get the rate limiting part from? 100.100.100.100 is provided by the Tailscale daemon on your local device and pretty much all resolving is done locally based on the DNS rules you have configured for your tailnet. MagicDNS will resolve your tailnet clients locally based on the network map, other lookups will be forwarded to your LAN or exit-node DNS / split DNS servers or to your configured global DNS servers if you have the override enabled in admin panel.
It is only resolving for devices in the Tailnet. Kuma is checking they are all up, and this Ansible playbook is checking they have all their updates. I wouldn’t have thought that was an unusual arrangement - and it’s worked perfectly for about a year till about three weeks ago.
It’s not only resolving for devices in the Tailnet if you saw apt hitting MagicDNS resolvers though. You have something misconfigured.
If you’re not familiar with networking or how DNS resolvers work, I’d start by checking your current node connected to the Tailnet and determine where your DNS requests are coming from with
digornslookup, then check your local resolver configurations, and other devices as well. Make absolutely sure that you’ve got a split DNS config (meaning you’re not using the fqdn of nodes in the Tailnet and not shortnames), and that the custom domain is showing as a target for searching on restricted resolvers in your resolver configs. Then branch out and check your other nodes to make sure they have a similar configuration.
