Administrator Guide

Server Performance metrics and workflows
7 Telemetry Streaming with iDRAC9 Server -Telemetry Performance Reports | ID 457
In the example, the output compares the telemetry report queried from an Intel-based Dell EMC PowerEdge
server before and after the naming schema modification has been implemented. AMD-based Dell EMC
PowerEdge servers will differ when compared with Intel-based servers. While the System Usage report for
Intel-based servers will contain metrics for CPU Usage, Memory Usage, IO Usage, and Aggregate Usage, the
System Usage report for AMD-based servers will only contain CPU Usage metrics.
In addition, any Telemetry Streaming scripts that were leveraging the old "CUPS" report name will require
modifications to support the new "System Usage" report name after iDRAC9 4.40.00.00 releases.
2.2 Aggregation Metrics report metrics
Aggregation Metrics report provides selected aggregate values which are useful to derive downstream
insights. The metrics that are in this report are listed below:
• SystemAvgInletTempHour - Average Inlet temperature for the last hour
• SystemMaxInletTempHour - Peak Inlet temperature for the last hour
• SystemMaxPowerConsumption - Peak system power consumption
2.3 CPU Memory Metrics report metrics
CPU Memory Metrics, or CPUMemMetrics, report provides selected CPU memory metrics and utilization
figures. The metrics that are in the report are listed below:
• CPUC0ResidencyHigh - CPU C0 Residency high value. 32 MSBs of the 64-bit counter containing the
aggregate C0 residency count from all the cores.
• CPUC0ResidencyLow - CPU C0 Residency Low value. 32 LSBs of the 64-bit counter containing the
aggregate C0 residency count from all the cores.
• CUPSIIOBandwidthDMI - CUPS IIO Bandwidth DMI Metric value which indicates utilization counter of
the DMI Port.
• CUPSIIOBandwidthPort<port number> - CUPS IIO Bandwidth Port specific Metric. Utilization counter
of the integrated I/O Port on a per x16 port granularity
2.4 Triggers for Server Performance reports
Telemetry reports can be configured to emit automatically on the event of selected life cycle triggers at the
server. The reports are sent across all active HTTPs subscription or SSE streams. Following table outlines
this behavior for performance reports.
Performance report
Qualifying server life-cycle trigger
SystemUsage
TMPCpuCriticalTrigger, TMPCpuWarnTrigger,
TMPMemCriticalTrigger, TMPMemWarnTrigger
AggregationMetrics
PULL only
CPUMemMetrics
None
For help with setting up triggers, see the “Telemetry – Getting Started” available at dell.com/support/idrac.