k3s: CoreDNS resolves AAAA when dualstack is not enabled
- Cluster was created without Dual-stack support
- Dual-stack support cannot be added after setting up the cluster according to this
- The host-nodes have IPv6 enabled
- The hosts-nodes DNS – Cloudflare in my case – returns AAAA records
- Therefore DNS lookups within the cluster also return AAAA records
- AAAA targets cannot be reached from pods within in the cluster
- CoreDNS’s config-map cannot be persistently overridden like this:
template ANY AAAA {
rcode NOERROR
}
How can this be solved?
About this issue
- Original URL
- State: closed
- Created 2 years ago
- Comments: 18 (7 by maintainers)
It kinda sounds like the legacy wireguard setup was dropping some portion of udp DNS traffic.
That’s not an error that dig would throw. That is an error from kubectl when trying to run the
execcommand. Rather than running a bunch of exec commands in a loop (and overloading the apiserver with exec requests), you might run the actual dig command in a loop within the exec command:kubectl exec -i -t dnsutils -- sh -c 'while true; do dig +all a google.com aaaa google.com; sleep 1; done'