Setting Power and Resource Policies

Setting Power Limits and Telemetry

The Power and Resource Policies section provides tools to monitor and control platform power consumption through the BMC. These capabilities support data-center power planning, long-term energy analysis, and automated power management using Node Manager (NM)–based policies.

NM Power Policy

The NM Power Policy page allows administrators to define and manage Intel® Node Manager power policies. These rules help enforce platform-level or component-level power caps and trigger actions when power thresholds are exceeded.

Policy Table

The Policy Table lists all configured Node Manager policies and includes the following fields:

• Policy ID Unique identifier for each power policy.

• Domain ID The target hardware domain where the power cap applies:

  • Entire Platform

  • CPU

  • Memory

• Enable Indicates whether the policy is currently active.

• Shutdown Specifies whether the system should shut down when the configured power limit is exceeded.

• Log Event Records a log entry when the policy conditions are triggered.

• Power Limit (Watt) Defines the maximum allowable power consumption for the selected domain.

Add / Edit NM Policies

Administrators can configure or modify a Node Manager policy using the following fields:

• Enable / Disable Policy Activates or deactivates the selected NM power policy.

• Response Action Choose one or both of the following actions when power exceeds the threshold:

  • Shutdown

  • Log Event

• Domain ID Specifies where the policy is applied:

  • Entire Platform

  • CPU

  • Memory

• Power Limit (Watt) Defines the exact power cap for the selected hardware domain.

Usage Notes

• Power policies are commonly used to enforce rack-level or data center power caps.

• Logging-only mode is helpful for power monitoring without disrupting uptime.

• Multiple policies may be configured to target different domains or operational profiles.

Monitoring Power Telemetry

The Power Telemetry page provides detailed, register-level power measurement data for supported hardware components. This feature enables precise long-term analysis of power trends and efficiency behavior.

Main Components

• Device Category Selector Selects the subsystem for which telemetry readings should be displayed (e.g., CPU, Memory, Platform components).

• Register Index The numeric index of the telemetry register being accessed.

• Register Address Hardware address of the telemetry register within the monitoring subsystem.

• Energy Counter (MJ) The accumulated energy consumption for the selected device, expressed in megajoules.

• Time Stamp (ms) The time (in milliseconds) when the telemetry reading was recorded.

Usage Notes

• If no data appears, the selected device category may not support telemetry or the feature may be disabled by firmware.

• Telemetry readings allow trend analysis, cross-load comparison, and identification of abnormal component power usage.

• Power telemetry is often used together with Power Statistics for real-time correlation and with NM Power Policy to define power caps based on historical data.

Reviewing Power Statistics

Power Statistics

The Power Statistics page provides a real-time overview of the system’s power consumption across major hardware domains. This allows administrators to track energy usage patterns, detect anomalies, and support data-center power optimization efforts.

Display Sections

The page presents power consumption data grouped by hardware domain:

• Entire Platform Shows total power consumption for the entire system, including all subsystems and components.

• CPU Displays real-time power usage for the processor(s).

• Memory Indicates the current power draw from system memory modules.

Data Indicators

Each domain includes the following measurement values:

• Instantaneous Power (Watt) Real-time power consumption at the moment of sampling.

• AVG (Average Power) Average power consumption calculated over the selected period.

• Period The time window used for average-power calculation.

Usage Notes

• If any reading displays N/A, the system may not support power monitoring for that domain, or sensors may be disabled or unavailable.

• Regular monitoring of power statistics can help identify unusual thermal or workload patterns and improve power efficiency planning.

• When used together with NM Power Policy, administrators can enforce power capping strategies based on observed power behavior.

Scheduling Power Events

Power Schedule

The Power Schedule feature allows administrators to configure automated power-on, power-off, or restart events based on specific dates, times, or recurring patterns. Scheduled operations help reduce manual intervention, optimize energy usage, and ensure systems are active only when required.

Interface Components

Each scheduling entry contains the following fields:

• Index Displays the entry number of the schedule rule.

• Enabled Indicates whether the schedule rule is Enabled or Disabled.

• Start Date / End Date Defines the date range during which the schedule is valid.

• Trigger Date Specifies the recurrence pattern, such as:

  • Daily

  • Weekly

  • Specific Date

• Trigger Time Sets the exact time when the scheduled action will execute.

• Action Defines the operation to be performed at the scheduled time:

  • Power On

  • Power Off

  • Restart

• Edit / Delete Allows modification or removal of an existing schedule entry.

• Add Creates a new power schedule rule.

• Delete All Removes all existing schedule entries.

Usage Notes

• A maximum of ten scheduled entries can be created.

• Ensure the system clock is configured correctly to maintain schedule accuracy.

• If multiple rules overlap, the most recently saved configuration takes precedence.

• This feature is useful for:

  • Lab or test environments

  • Production facilities with defined operating hours

  • Energy-saving policies where systems do not need to run continuously