UCS - Unified Computing System

Overview

Cisco UCS standalone/CIMC and UCS Manager provide the physical compute and management layer hosting UC and support workloads.

Monitoring

Method

SNMP and API collection through ucs-exporter and infra-prometheus.

Metrics Collected

• Chassis, server, and service-profile health
• Power, hardware, interface, and fault indicators
• CPU, memory, and environmental status
• Exporter and scrape health

Grafana Dashboards

• WxCDI UCS SA Detail and UCSM Detail are expected by the audit; verified dashboard links are not yet recorded.

Logging

Log Source

SNMP traps are expected. The audit still lists the UCS/UCSM syslog index as undefined.

Splunk Index

snmp_traps; verify the current syslog destination before searching.

Key Log Queries

index=snmp_traps device_type IN ("ucs_sa","ucsm") device_name="<device_name>"

Alerting

Grafana Alert Rules

Alert Name	Severity	Description
UCS hardware health	Varies	Compute and hardware fault conditions.
Scrape-down coverage	Gap	The audit currently records this control as disabled.

Splunk Alerts

Alert Name	Severity	Description
UCS event coverage	Gap	Complete and verify the syslog/index mapping.

Notification Flow

UCS → Grafana/Splunk → EMS (KeepHQ) → [notification channels]

Inventory

• Repository: ucs-sa-inventory and ucsm-inventory
• Update Method: UCS standalone is maintained manually; UCSM is generated with documented manual exceptions. Exporter target config may also require a wxcdi-kubed-configs change.
• Update Cadence: On UCS/UCSM inventory change.