Management, Monitoring, and Automation
Domain 6 - Management and Automation
This domain tests observability, auditability and repeatable operations.
Core rules
- Cloud Insight collects metrics and renders dashboards; Event Rules turn thresholds into notifications or actions. Hypervisor-level metrics (CPU, network, disk I/O) are available by default; memory and filesystem usage require the monitoring agent on the server.
- Cloud Activity Tracer records control-plane API calls per account - who changed what, when. Export logs if you need retention beyond the built-in window.
- Automation surfaces: Ncloud CLI, REST APIs and SDKs, the DevOps toolchain (SourceCommit, SourceBuild, SourceDeploy, SourcePipeline) for CI/CD, and Cloud Functions for event-driven code.
- Infrastructure as code is expected: script creation of VPC, subnets, ACG, servers and load balancers so environments are reproducible.
Worked example
Autoscale a web tier when CPU stays above 70% for five minutes. Reasoning: create a Launch Configuration (image, spec, init script, ACG), an Auto Scaling Group bound to the subnets, and a scale-out policy with a cooldown so it does not thrash; add a Cloud Insight Event Rule on CPU to notify on-call. Scale-in needs a longer threshold than scale-out because removing capacity is riskier.
Traps
- Alerting on a single one-minute spike with no duration.
- Expecting memory metrics without installing the agent.
- Editing production by hand, then wondering why the next pipeline run reverts it.
- No cooldown, so the group oscillates between minimum and maximum.