Thermal Management

Introduction

Thermal Management (Thermal Control) is an array-wide cooling group control system designed by aBMC for array servers. The system consists of two major modules: status monitoring and temperature control.

Monitoring module: collects hardware thermal information such as SOC temperatures across the array and real-time fan speeds.
Temperature control module: provides three preset cooling strategies—Silent, Balanced (default), and Powerful—along with customizable temperature control settings, so users can choose flexibly based on data center environment and compute load.

Version Requirement

aBMC version >= v1.3.0

Development Vision

Provide users with visualized temperature charts, making it easy for customers to adjust algorithm intensity based on cooling conditions.
The Balanced cooling strategy of the temperature control module helps effectively prevent single-point thermal runaway in array clusters.

Feature Usage

Setting the Fan Operating Mode

Open the fan settings window

Select System in the left navigation bar.
Select Thermal in the secondary navigation bar.
On the Fan Info page, check the fan presence status under Summary.
Check Status and RPM under Details.
Click Setting to open the Fan Settings window.

Select the fan operating mode

In Mode, select Silent, Balanced, Powerful, or Custom.
If Custom is selected, enter an integer between 1–20 in Gear Level.
After verifying the configuration, click Confirm.

Confirm the configuration result

Return to Fan Info, confirm that Speed Adjustment Mode has been updated, and continue to observe whether RPM and board temperatures are as expected.

Mode selection guidelines

Mode	Description	Recommendation
Silent	Prioritizes reducing fan noise.	Suitable for environments with low load and stable temperatures. Continue monitoring temperatures after applying.
Balanced	Balances cooling capacity, power consumption, and noise.	Recommended as the common mode for typical operating scenarios.
Powerful	Increases cooling capacity.	Suitable for high-load or high-temperature scenarios; noise and power consumption may increase.
Custom	Controls fans using the specified Gear Level.	Use only when you understand the device cooling requirements; verify temperature and RPM after configuration.

About Custom mode

The valid range for Gear Level is 1–20. The mapping between gear levels and actual RPM or cooling performance is determined by the device, and the gear value does not represent a speed percentage. After making changes, continue to monitor fan speed and board temperatures.

View fan information

When --mode is not specified, the command returns current fan and temperature information.

bmc fan --protocol http --ip <BMC_IP> --port <BMC_PORT> --user <USERNAME> --password <PASSWORD>

Set a preset operating mode

Choose one of the preset modes to execute, based on cooling and noise requirements.

Silent mode

bmc fan --protocol http --ip <BMC_IP> --port <BMC_PORT> --user <USERNAME> --password <PASSWORD> --mode Silent

Balanced mode

bmc fan --protocol http --ip <BMC_IP> --port <BMC_PORT> --user <USERNAME> --password <PASSWORD> --mode Balanced

Powerful mode

bmc fan --protocol http --ip <BMC_IP> --port <BMC_PORT> --user <USERNAME> --password <PASSWORD> --mode Powerful

Set a custom gear level

When using Custom mode, you must specify a gear level between 1–20 via --level.

bmc fan --protocol http --ip <BMC_IP> --port <BMC_PORT> --user <USERNAME> --password <PASSWORD> --mode Custom --level 20

Demo

The following example assumes a BMC address of 172.16.100.173, service port 443, and username admin. Replace admin with the actual password, and adjust the address, port, and account according to your deployment.

# Set to Balanced mode
bmc fan --protocol http --ip 172.16.100.173 --port 2108 --user admin --password admin --mode Balanced

# Query the configuration result
bmc fan --protocol http --ip 172.16.100.173 --port 2108 --user admin --password admin

Parameter description

Parameter	Required	Description
`--protocol`	Yes	Request protocol, for example `http`.
`--ip`	Yes	BMC management address.
`--port`	Yes	BMC service port.
`--user`	Yes	HTTP Basic authentication username.
`--password`	Yes	HTTP Basic authentication password.
`--mode`	Required when setting a mode	Valid values are `Silent`, `Balanced`, `Powerful`, or `Custom`.
`--level`	Required for Custom mode	Fan gear level, in the range `1–20`.
`--output-format`	No	Specifies the client output format.

Credential security

Command-line arguments may be retained in shell history or process lists. Avoid using real passwords directly in shared environments, and use a secure credential delivery method appropriate for your deployment.

Operation	Method	URI
Query action parameters	GET	`/redfish/v1/Chassis/bmc/FanModeActionInfo`
View fan and temperature information	GET	`/redfish/v1/Chassis/bmc/Thermal`
Set the fan operating mode	POST	`/redfish/v1/Chassis/bmc/Actions/FanMode.Setting`

Note

For detailed information about interface authentication, request parameters, response fields, and error codes, refer to the Redfish API documentation.

Viewing Thermal Monitoring Charts

Open the temperature monitoring page

Select System > Thermal in the left navigation bar.
Click Ocean Temp at the top of the page.
Wait for the device topology, temperature data, and trend charts to finish loading.

View the temperature topology

Check the current temperature and status indicators of each monitoring point.
To adjust the view, use the topology controls to zoom, fit, or reset.

Configure temperature trend parameters

Use Select Time Range to choose the historical data range, for example Last 30 Min.
Set the data sampling Step within the recommended range displayed beside the input box.
Select the automatic refresh interval, for example 15 Seconds.
To update the data immediately, click Refresh.

Confirm the monitoring result

Confirm that the temperature monitoring points in the device topology are displayed correctly.
Confirm that the time range and data density of the trend chart meet expectations.
After a manual refresh or waiting for the automatic refresh, confirm that temperature data and status update properly.
If a monitoring point shows an abnormal status or the temperature keeps rising, perform further troubleshooting in combination with fan RPM, the current operating mode, and device alarms.

Temperature monitoring parameter description

Parameter or area	Description	Usage requirements
Device Topology	Displays boards, interfaces, temperature values, and status points according to the physical device layout.	Judge whether temperatures are normal in combination with device specifications, current load, and device alarms.
Select Time Range	Selects the query time range for temperature trend data.	For larger time ranges, a larger sampling step is recommended.
Step	Controls the sampling interval of chart data. Smaller values produce denser data points; larger values produce sparser data points.	Must stay within the system-recommended range displayed beside the input box.
Refresh Interval	Controls how often the page automatically fetches the latest data.	Shorter intervals result in more frequent requests; choose based on monitoring needs.
Refresh	Immediately fetches temperature and status data again.	Use a manual refresh when data has not updated for a long time.
Topology Controls	Used to zoom, fit the view, or reset the topology.	Affects the page view only and does not change device state.

About the sampling step

Step controls the data sampling interval of the trend chart. Smaller values produce denser data points; larger values produce sparser data points. Use the recommended range provided by the page for the current time range, and avoid excessively dense data points that could affect query and rendering performance.