Lightning Talk: Is Your Kubernetes Cluster's DNS Working?
Lightningat KubeCon + CloudNativeCon North America 2019 · on Tuesday, 2019-11-19· at 18:02· Exhibit Hall AB - San Diego Convention Center, Ground Level·
summary
Kubernetes DNS failures can come from kernel and libc behaviors not visible to CoreDNS metrics: a kernel DNAT race drops parallel DNS packets and removed-endpoint state leaves queries mapped to dead replicas, and musl parallel queries plus strict DNS handling breaks some cloud resolvers. Measuring UDP sockets and correlating to pods yields per-service timeouts, latencies and volumes; using eBPF per-service DNS metrics produces dashboards that reveal misconfigured services and capacity shortfalls in practice.
AI-written summary · about
chapters (7)
- 0:00Intro
- 1:25Parallel DNS queries
- 2:01Linux Kernel LibC
- 2:59Core DNS Monitoring
- 3:32Application Visibility
- 4:31Core DNS Visibility
- 5:02Other Use Cases
related talks
28mHow to Break your Kubernetes Cluster with Networking
35mKubernetes, Lessons Learnt
32mDemystifying DNS: A Guide to Understanding and Debugging Request Flows in Kubernetes Clusters
32mUnderstand and Troubleshoot the “Magic” of Kubernetes Networking - Minhan Xia & Rohit Ramkumar
34mIt’s Always DNS: So Let’s Break It On Purpose
38mLogs Told Us It Was DNS, It Looked like DNS, It Had to Be DNS, It Wasn't DNS
03mLightning Talk - Introducing CoreDNS Kubernetes Integration
29m“Are These Things Talking to Each Other?” - Observing Kubernetes Networking...