Dell PowerEdge R620 fans are constantly at +20% spin when ConnectX-4 LX adapter is attached

Infrastructure & Networking Adapters and Cables
1

We have purchased a few MCX4121A-ACAT adapters recently for using on Dell PowerEdge R620 and they work well except one severe issue.

The server fans start spinning at 20% once OS start and never stop after (screenshot included). Normally the fans work at 1% while system is on idle (screenshot included). The issue is persistent on any OS both in UEFI and BIOS boot modes and is more likely related to the firmware of the adapter itself.

We have tried setting BIOS settings conformed to settings from here https://community.mellanox.com/s/article/bios-performance-tuning-example-for-dell-poweredge-r730 but there is the same +20% fan spin increase at the OS boot and persistently after.

The firmware upgrade didn’t fix the issue.

The following is the output of mlxup tool that shows current firmware version and other information:

./mlxup

Querying Mellanox devices firmware …

Device #1:

Device Type: ConnectX4LX

Part Number: MCX4121A-ACA_Ax

Description: ConnectX-4 Lx EN network interface card; 25GbE dual-port SFP28; PCIe3.0 x8; ROHS R6

PSID: MT_2420110034

PCI Device Name: 0000:04:00.0

Base MAC: 0c42a14a1964

Versions: Current Available

FW 14.28.1300 14.28.1300

PXE 3.6.0101 3.6.0101

UEFI 14.21.0016 14.21.0016

Status: Up to date

I would like to know if there is any possible solution for this.

with_mellanoxCX4LX.png
with_mellanoxCX4LX.png1290×560 121 KB

without_mellanoxCX4LX.png
without_mellanoxCX4LX.png1307×567 125 KB

2

Hello Lilian,

Thank you for posting your inquiry on the NVIDIA Networking Community.

Based on the information provided, we recommend to update the BIOS of the server to the latest version available which is version 2.9.0 and reset all settings to factory default. If the issue still persist, please open a support case with Dell Support. They can provide you guidance on how-to adjust the default fan-speed offset.

Basic when adding more components to a system, some manufactures will automatically up-en the fan speed to provide adequate cooling to all components in the system.

Thank you and regards,

~NVIDIA Networking Technical Support

3

Thank you, but the BIOS version is already at 2.9.0 and doesn’t actually fix anything but Intel vulnerabilities.

While it’s true that sometimes server may add cooling for additional components, it’s not normal that server does it when it’s not needed, because in our case it happens on idle without any activity or cables attached.

We have already opened an issue with Dell and they told us that it happens due to improper thermal settings in the PCI device’s firmware.