22_check_k8s_warning_events.yml - Usage Guide

Read-Only k3s kubeadm

Purpose

Collects Kubernetes events of type Warning in the cluster from the first control-plane:

  • Last N hours (default 24)
  • Reason frequency
  • Namespace distribution
  • Recent event lines (time, count, kind/name, message)

Read-only.

Variables

Variable Default Description
events_since_hours 24 How far back to look
events_limit 80 Max events to list
events_namespace "" (empty) If set, only that namespace; if empty, all namespaces 
# All namespaces
ansible-playbook -i inventories/cagatayuresincom/hosts.ini playbooks/22_check_k8s_warning_events.yml \
  --extra-vars 'events_since_hours=48 events_limit=120'

# A single namespace
ansible-playbook -i inventories/cagatayuresincom/hosts.ini playbooks/22_check_k8s_warning_events.yml \
  --extra-vars 'events_namespace=n8n'

How to run 

ansible-playbook -i inventories/cagatayuresincom/hosts.ini playbooks/22_check_k8s_warning_events.yml

Interpretation

Reason (e.g.) What to check
FailedScheduling Resources / taint / affinity
FailedMount / FailedAttachVolume PV/storage → 23
ImagePullBackOff / ErrImagePull Registry / image → 17, 19
OOMKilled / BackOff Limit / memory → 10, 15
Unhealthy Probe / application health → 01

Notes

  • Script: playbooks/files/k8s_warning_events_check.py
  • Events are stored in etcd and expire over time; “empty for the last 24h” does not always mean there is no problem.