troubleshoot-kubernetes-kube-proxy

What problem does it solve?

Troubleshoots Kubernetes Kube Proxy incidents where service traffic breaks or networking behavior degrades without obvious application-layer errors, such as stale rules, conntrack exhaustion, sync loop storms, iptables lock starvation, or silent conntrack leaks.

Core Features & Use Cases

• MCP-driven triage against Netdata signals: Uses Netdata MCP to query Kubernetes Kube Proxy health signals and narrows the failure archetype using the operator playbook’s diagnostic tree.
• Failure-pattern guided diagnosis: Focuses on key kube-proxy control-plane failure modes (stale rules/zombie watch, conntrack table full, sync loop storm, iptables lock starvation, silent conntrack leak, and lost changes on restore failures).
• Verification-first workflow: Recommends remediation routes and instructs how to re-query the same MCP contexts to confirm the system returned to an expected range.

Quick Start

Use this skill when diagnosing Kube Proxy anomalies by asking your agent to troubleshoot Kube Proxy for stale rules or conntrack issues using Netdata MCP and the operator playbook diagnostic tree.