troubleshoot-coredns

What problem does it solve?

It helps diagnose and localize CoreDNS outages by mapping common failure archetypes (upstream black holes, cache collapse, Kubernetes API disconnect, memory blowouts, and forwarding loops) to the Netdata-observed health signals that reflect root cause.

Core Features & Use Cases

• Failure-mode triage tree for CoreDNS: Routes an on-call or engineering agent through a structured diagnostic path tailored to distinct CoreDNS failure archetypes rather than generic “is the pod up” checks.
• Netdata MCP-driven signal verification: Uses MCP to query CoreDNS-specific Netdata contexts, pull recent anomalous metrics, and validate findings against load-bearing chart contexts.
• Remediation confirmation loop: Recommends remediation and then instructs re-querying the same MCP verification set to confirm signals return to expected bands.
• Use case: You get paged for elevated DNS errors/latency or CrashLoopBackOff in Kubernetes; this skill walks you through identifying whether the underlying cause matches upstream reachability, caching behavior, Kubernetes API connectivity, memory/OOM, or misconfigured forwarding loops.

Quick Start

Use the troubleshoot-coredns skill to diagnose why my CoreDNS service is failing (upstream unreachable, cache collapse, kubernetes api disconnect, memory blowout, or forwarding loop) and then recommend the remediation based on Netdata MCP health signals.