Health Monitoring, Troubleshooting, and Support Escalation
This section covers Health Monitoring, Troubleshooting, and Support Escalation: NCC checks, alert triage, log collection, Pulse, support case preparation, and the difference between local remediation and vendor escalation.
Monitoring and escalation rules
- NCC runs health checks across AOS, AHV, storage, networking, and cluster services. Critical failures should be addressed before upgrades or expansion.
- Triage by scope: single VM, single host, whole cluster, or management plane. Scope narrows the likely cause.
- Collect logs and NCC output before destructive actions such as rebooting a CVM or node.
- Pulse sends diagnostic information to Nutanix support when enabled; it does not replace a support case.
- Escalate with cluster ID, case severity, timeline, symptoms, recent changes, and collected logs.
Worked example
A cluster reports a CVM unreachable alert. First check whether one CVM or all CVMs are affected. If only one is down, verify the host, network, and CVM service state; collect logs before rebooting. If all CVMs are unreachable, suspect management network or upstream switching. Rebooting every CVM immediately would remove evidence and may worsen an outage, so escalate after preserving diagnostics.
Traps that cost marks
Do not open a case without basic scope and logs. Do not confuse Pulse with the support portal. Do not reboot before collecting diagnostics when the cluster is degraded. Do not ignore NCC failures because the UI looks healthy.