Cloud Operations, Monitoring, and Lifecycle Management
This domain tests how a cluster is watched and how it is upgraded without breaking itself: health checking, alerting, inventory, and upgrade sequencing.
Monitoring surfaces
- Prism Element and Prism Central dashboards surface alerts, capacity, and performance.
- NCC (Nutanix Cluster Check) runs health checks on a schedule and on demand; it is the first stop when support asks for diagnostics.
- Insights aggregates findings across clusters; Pulse sends telemetry to Nutanix and is opt-in.
- Audit logs and syslog forwarding provide the evidentiary trail.
Lifecycle management
- LCM inventories software and firmware, flags prerequisites, and performs rolling one-click upgrades with no full outage.
- Upgrading Prism Central ahead of the clusters it manages avoids a management plane that cannot understand newer cluster versions.
- Disconnected sites need dark-site configuration with pre-downloaded bundles.
- Foundation images new nodes and supports cluster expansion.
Worked example
Before an AOS upgrade: run NCC to confirm no blocking checks, run an LCM inventory so prerequisites and firmware pairings are current, upgrade Prism Central first, then roll the AOS upgrade across nodes one at a time, and schedule firmware separately. Each step has an exit criterion - a failed NCC check or unmet prerequisite stops the sequence rather than being noted and ignored.
Traps
Treating NCC output as equivalent to live monitoring alerts; skipping the Prism Central upgrade; assuming LCM works without internet or dark-site bundles; expecting an AOS upgrade to require a full cluster outage.