skills.rest
schoolofdevops
schoolofdevopsOFFICIAL
Explore all skills by this author →
WebsiteGitHub

sre-k8s-node-health

Diagnose Kubernetes node health conditions using read-only kubectl commands.

Updated Apr 5, 2026
Search Tags:#kubernetes#diagnosis#kubectl#sre#memory pressure#node health#notready
One-click install
npx skills add https://github.com/schoolofdevops/401 --skill sre-k8s-node-health-schoolofdevops
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sre-k8s-node-health
Source: https://github.com/schoolofdevops/401/tree/main/skills/sre-k8s-node-health
Command: npx skills add https://github.com/schoolofdevops/401 --skill sre-k8s-node-health-schoolofdevops

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a Kubernetes node enters NotReady, MemoryPressure, DiskPressure, or PIDPressure, on-call engineers need a safe, read-only diagnostic procedure that gathers the right evidence without risking destructive actions like drain or cordon. ## Core Features & Use Cases - Node Condition Diagnosis: Collects node inventory JSON, describe output, resource usage, and per-node pod lists via four deterministic kubectl steps. - Pressure Analysis: Interprets Ready, MemoryPressure, DiskPressure, and PIDPressure conditions against allocatable capacity and pod resource requests. - Safe Escalation: Enforces read-only operation with explicit NEVER DO rules and structured escalation handoffs including full command outputs. - Use Case: A node shows MemoryPressure and pods keep getting evicted. Run this skill to gather node conditions, identify the highest-memory pods as eviction candidates, and escalate with complete evidence—without executing any mutating commands. ## Quick Start Ask the AI to diagnose why a Kubernetes node is in NotReady or MemoryPressure state using read-only kubectl commands.

Frequently Asked Questions about sre-k8s-node-health

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I diagnose a Kubernetes node in NotReady state?▼

Run kubectl get nodes -o json to inspect status.conditions, then kubectl describe node to review the Events section for kubelet failures. Check whether the Ready condition's message indicates a kubelet outage, network partition, or node OOM before escalating.

How to troubleshoot MemoryPressure on a Kubernetes node?▼

Check the node's MemoryPressure condition, then compare allocatable.memory from kubectl describe node against the summed memory requests of all pods on that node. Identify the highest-memory pods as eviction candidates and escalate rather than deleting pods directly.

Does this skill support offline or mock cluster testing?▼

Yes, setting HERMES_LAB_MODE=mock switches to a mock-kubectl wrapper that reads pre-baked node JSON fixtures instead of a live cluster. Add the course/infrastructure/wrappers/ directory to PATH to enable mock mode.

Can I use this skill to fix a pressured node automatically?▼

No, this skill is strictly read-only diagnosis. It never executes kubectl drain, cordon, delete node, or SSH into nodes; remediation actions require explicit approval from an on-call engineer through a separate procedure.

What kubectl permissions are needed for node health diagnosis?▼

You need read-only cluster access: get and list on nodes, get and list on pods across all namespaces, and describe node. No write or delete permissions are required since the skill performs no mutations.

Related Skills

Explore More Software Engineering Skills→
3horizons3horizons

kubectl-cli

Run kubectl commands for Kubernetes cluster management and troubleshooting.

Official
Intermediate
qa-amanqa-aman
19

runbook

Create operational runbooks with health checks, escalation paths, and recovery steps.

Community
Advanced
alon3153alon3153

gke-workload-scaling

Configures HPA and VPA autoscaling for GKE workloads using kubectl and YAML manifests.

Community
Intermediate
gfargogfargo

monitoring-setup

Deploy Prometheus, Grafana, and Alertmanager for self-hosted monitoring across VPS stacks.

Community
Advanced