Operations, Monitoring, and Lifecycle Management
Day-2 tooling
Prism Element manages one cluster; Prism Central manages many. NCC (Nutanix Cluster Check) runs health checks; LCM (Life Cycle Manager) plans and executes software and firmware updates. The NC2 console handles cluster-level lifecycle: deploy, scale up or down, upgrade orchestration, and delete. AWS CloudWatch and VPC Flow Logs cover the infrastructure layer.
Monitoring rules
- Configure alert email, SNMP, and syslog early; Prism alerts are only useful if they reach someone.
- Watch capacity separately from performance: local NVMe consumption, snapshot usage, and replication backlog.
- Use NCC before and after changes; a clean NCC baseline makes upgrade failures easier to diagnose.
- Track AWS quotas and bare-metal availability; a healthy Nutanix cluster can still fail to scale because AWS has no capacity.
Worked example
After adding a node, Prism shows a health warning and NCC reports a CVM discrepancy. The operator checks the NC2 console task, confirms the node joined, then runs NCC and LCM inventory. LCM shows an AOS version mismatch between the new node and the cluster. The fix is to bring the new node to the cluster's AOS version through the supported LCM/NC2 workflow, not to manually patch the CVM.
Traps
Upgrading AOS by hand instead of LCM; deleting EC2 instances directly; ignoring NCC warnings until an upgrade fails; assuming Prism Central is mandatory for a single cluster.