Add the ability to disable the third-party PCIe card default cooling

[tigerblue77]

• maybe apply only if parameter supplied as boolean environment variable ?
• maybe apply only if PCIe card present ?

Enable third-party PCIe card default cooling response

ipmitool -I lanplus -H <IPADDRESS> -U <USERNAME> -P <PASSWORD> raw 0x30 0xce 0x00 0x16 0x05 0x00 0x00 0x00 0x050x00 0x00 0x00 0x00

Disable third-party PCIe card default cooling response

ipmitool -I lanplus -H <IPADDRESS> -U <USERNAME> -P <PASSWORD> raw 0x30 0xce 0x00 0x16 0x05 0x00 0x00 0x00 0x05 0x00 0x01 0x00 0x00

Get Third-Party PCIe Card Default Cooling Response Logic Status

ipmitool -I lanplus -H <IPADDRESS> -U <USERNAME> -P <PASSWORD> raw 0x30 0xce 0x01 0x16 0x05 0x00 0x00 0x00

The response data is:

• 16 05 00 00 00 05 00 01 00 00 (Disabled)
• 16 05 00 00 00 05 00 00 00 00 (Enabled)

Source : https://www.dell.com/support/kbdoc/en-us/000135682/how-to-disable-the-third-party-pcie-card-default-cooling-response-on-poweredge-13g-servers

Other links :

• https://manualzz.com/doc/7482018/disabling-a-third-party-pcie-card-cooling
• https://fohdeesha.com/docs/store/perc/ThirdPartyPCIFanResponse.pdf

[louispires]

This would be great!!

As when my server goes over the CPU_TEMPERATURE_TRESHOLD, it then jumps to 10k RPM instead of something more applicable to 55 Degrees.

[tigerblue77]

Hello @louispires, thanks for your interest in this project. I don't have much time to devote to this project at the beginning of the year (some difficulties on personal non-IT projects).

The problem that you describe is the actual "normal" behavior and is not linked to "Third-Party PCIe Card default cooling response" functionality, if I can call it like this. When the defined CPU_TEMPERATURE_TRESHOLD is reached, this Docker container immediately resets the server's default fan mode for safety, in order to avoid overheating (or at least to prevent the CPUs from exceeding this threshold temperature too much). There is no algorithm to progressively increase the fan speed in this container.

What I can advise is :

• to check Tcase (case temperature) of your CPU on Intel Ark website and then set CPU_TEMPERATURE_TRESHOLD to a slightly lower value. Example with my CPUs (Intel Xeon E5-2630L v2) : Tcase = 63°C, I set CPU_TEMPERATURE_TRESHOLD to 60(°C).
• If it's already good, adapt your FAN_SPEED value to increase the airflow and thus further decrease the temperature of your CPU(s)
• If neither increasing the fan speed nor increasing the threshold solves your problem, then it may be time to replace your thermal paste?

[louispires]

Hey @tigerblue77,

Sorry my initial explanation might have been wrong, the project works great and exactly as I would expect it to.

Due to my server containing a NON Dell approved PCI-E Card, technically the default Fan Curve is much higher, as you explained when opening this issue.

Thus, when defaulting to the Dell Fan Curve, it jumps to 10k rpm, instead for the temperatures, it should be closer to 6k:

Here is the exact same settings, but with the PR #33 in place: