The Linux kernel's UDP connection tracking issues have caused challenges with DNS, particularly in Kubernetes, where elevated rates of DNS failures were observed under load. A race condition in the kernel causes the loss of response for one or two queries when made in a short period. To address this issue, a new implementation called coreDNS-codecache was developed as a CoreDNS plugin to solve some of the new problems introduced by node-local, which is a Kubernetes extension that solves these issues but also introduces others. The new plugin uses a DNS cache on each node, greatly reducing outgoing DNS queries and upgrading requests to TCP. However, it brings its own set of concerns, including high availability issues and limitations with Alpine Linux containers. To address these concerns, the developers created Coredns-nodecache, a plugin that supports setup in high availability and has been used in production for several months. The solution is still evolving, with ongoing efforts to improve it by delegating the creation of iptables rules and interface to a Kubernetes operator.