CoreDNS 1.7: Resilient DNS for Growing Clusters

Introduction

CoreDNS 1.7, released on June 15, 2020, focuses on resiliency under heavy load. Negative caching, configuration reloads, and TCP handling all receive attention to keep Kubernetes DNS fast and predictable.

Smarter Caching and Timeouts

• Negative caching fixes prevent repetitive lookups for nonexistent services, slashing control-plane chatter during rollouts.
• TTL handling honors upstream responses precisely, improving compatibility with hybrid DNS setups.
• Timeouts gained more granular configuration, letting operators tailor query and connect windows per environment.

Graceful Reloads & Observability

• Hot reloads no longer drop in-flight queries; CoreDNS now waits for active connections to drain before applying config changes.
• Enhanced logging includes structured fields and per-plugin verbosity levels.
• Prometheus metrics add histogram buckets for request latency, aiding SLO tracking.

TCP & Edge Cases Hardening

1. TCP Fast Open support improves performance for workloads with frequent DNS-over-TCP usage.
2. Concurrent request limits guard against noisy neighbors and DNS amplification attempts.
3. Bug fixes in the forward and grpc plugins handle upstream failures more gracefully.

Getting Started

kubectl set image -n kube-system deployment/coredns \
  coredns=coredns/coredns:1.7.0
kubectl rollout status -n kube-system deployment/coredns

Monitor metrics:

kubectl port-forward -n kube-system svc/coredns 9153:9153

Summary

CoreDNS 1.7 ensures the DNS plane scales alongside application traffic, giving teams confidence to grow without outages.