[Linux] Dell PowerEdge Servers Run Fans at Maximum Speed Due to PCIe Thermal Policy

[Linux] Dell PowerEdge Servers Run Fans at Maximum Speed Due to PCIe Thermal Policy

Environment



Item
Details
Server PlatformDell PowerEdge  XE7745
RAID CardSupremeRAID™ SR1001, SR1000, SR1010
GPUNVIDIA T400, NVIDIA T1000, NVIDIA A2000 
Management InterfaceDell iDRAC
Operating SystemLinux (distribution independent)



Issue / Symptom

After installing SupremeRAID™ SR1001 (NVIDIA T400) into a Dell PowerEdge server, the system fans remain at maximum speed (100% PWM) continuously.

This behavior persists even when:

  • The system is idle

  • CPU and GPU utilization is low

The excessive fan speed results in unexpected noise levels and may be mistaken for a hardware or driver issue.


Root Cause

This behavior is expected behavior controlled by Dell iDRAC thermal policy.

When iDRAC detects a third-party PCIe device that is not listed in Dell’s internal thermal database, it cannot determine the correct cooling profile for that device.

To prevent potential overheating, iDRAC automatically:

  • Enables High LFM (Linear Flow per Minute) mode

  • Forces the associated fan zone to maximum speed

The SupremeRAID™ SR1001 (T400) is detected as a non-Dell PCIe device, which triggers this safety mechanism.


Resolution / Solution

You can manually override this behavior by disabling LFM mode for the specific PCIe slot where the SR1001 (T400) is installed.

This is done using the racadm command via iDRAC.


Procedure

Step 1 – Connect to iDRAC via SSH

ssh root@<iDRAC_IP>

Step 2 – Identify the PCIe Slot with LFM Enabled

List the LFM configuration for all PCIe slots:

racadm get system.PCIeSlotLFM

Example output:


If the SR1001(T400) is installed slot 1

Step 3 – Disable LFM Mode for the Target Slot

Set LFMMode to 1 for the identified slot:

racadm set system.PCIeSlotLFM.1.lfmmode 1
  • 1 = Custom / Disabled LFM mode

  • This change applies only to the specified PCIe slot


Step 4 – Verify the Configuration

Confirm the setting:

racadm get system.PCIeSlotLFM.1

Once applied, the system fans should gradually return to normal operating speed within a short period.


Important Notes / Warnings

⚠️ Thermal Safety Notice

By disabling LFM mode:

  • You are bypassing Dell’s fail-safe cooling policy for this PCIe slot

  • Thermal responsibility for the GPU shifts to the OS level


Reference Link:

Dell iDRAC10 System Attributes Guide — PCIe Slot LFM Configuration
This Dell manual explains system thermal attributes including PCIe slot LFM (Linear Flow per Minute) control settings in iDRAC.