CheckMK Troubleshooting Blueprint
Systematic approach to diagnosing and resolving common CheckMK issues — agent connectivity, check failures, notification problems, and performance.
Difficulty
Intermediate
Estimated time
2–3 hours
Tags
What this helps with
- ✓Agent connectivity issues (CRIT/UNKNOWN)
- ✓Service discovery problems
- ✓Notification pipelines not firing
- ✓High check latency
- ✓Rule conflicts and overrides
Step-by-step plan
- 1
Identify the problem scope
Single host, single service, or widespread? Check the main dashboard for patterns. Use Analyze → Audit log for recent changes.
- 2
Check agent connectivity
Run cmk -d <hostname> to fetch agent data manually. Check TCP 6556 is open. Verify the agent is running on the target host.
- 3
Re-run service discovery
Navigate to the host → Service configuration. Run "Fix all" or manually accept discovered services. Activate changes after.
- 4
Trace notifications
Use Notifications → Notification log. Check the notification rule order. Test with "Fake check results" to trigger a notification manually.
- 5
Check performance metrics
Review cmk --debug -n <host> for slow checks. Increase check intervals for noisy services. Consider a distributed setup for large environments.
Download this blueprint
CheckMK Troubleshooting Checklist (PDF)