15_get_metrics_server_stats.yml - Usage Guide

Purpose

If metrics-server is ready, reports node and pod CPU/memory usage via kubectl top in a readable form. If it is not installed, it does not install; it prints a warning (use 16 to install).

Shared tasks:

• tasks/metrics_server_check.yml
• tasks/metrics_server_report.yml

Report sections

Section	Command	How to read
Node	kubectl top nodes	Instant CPU (cores) and MEMORY per node
Pod CPU	kubectl top pods -A --sort-by=cpu	Highest CPU pods first
Pod memory	kubectl top pods -A --sort-by=memory	Highest RAM pods first

Values are instantaneous (metrics-server scrapes ~15s); this is not a long-term graph. Interpret high usage together with 10 / 11.

Requirements

• master / singlenode, first control-plane
• metrics-server must be READY (Available=True). If not, use 14 or 16.

How to run

ansible-playbook -i inventories/cagatayuresincom/hosts.ini playbooks/15_get_metrics_server_stats.yml

Notes

• Right after a fresh install you may see “no data”; waiting ~1 minute and re-running 15 is enough.
• Read-only.