cooling-failure-protection

Cooling Failure Protection for Servers/NAS/Desktops/etc.

Introduction

This Tool is meant to be a (almost) last Resort to prevent permanent Damage or Severe Lifespan Reduction of your Drives.

It was Developed after a Cooling Failure caused several of my HDDs to reach very high Temperatures.

For a Variable Fan Speed Control for Supermicro IPMI Motherboards, please refer to my Supermicro Fan Controller Repository.

Features

At the moment only Drives (HDD, SSD, NVME) are supported.

!! NVME Temperature Reading hasn't been extensively Tested !!

Requirements

This requires the following Tools to be Installed:

• jq
• smartctl

Installation

Clone the Repository:

git clone https://github.com/luckylinux/cooling-failure-protection.git

Change Folder to the Project that was just cloned:

cd cooling-failure-protection

Run the Setup:

./setup.sh

Configuration

General

Fahrenheit Value is only displayed for Reference, it is NOT used in ANY Part of the Code !

The following Configuration Variables are currently defined:

For Device	Variable	Default Value °C	Equivalent Value °F	Description	Action
HDD	hdd_warning_temp	45	113	Warning Temperature for HDDs	Beep
HDD	hdd_shutdown_temp	50	122	Shutdown Temperature for HDDs	Shutdown
SSD	ssd_warning_temp	60	140	Warning Temperature for SSDs	Beep
SSD	ssd_shutdown_temp	70	158	Shutdown Temperature for SSDs	Shutdown
NVME	nvme_warning_temp	70	158	Warning Temperature for NVMEs	Beep
NVME	nvme_shutdown_temp	80	176	Shutdown Temperature for NVMEs	Shutdown

Default Settings

The Default Settings are installed in /etc/cooling-failure-protection/default.sh.

User Settings

The User can define Custom Settings in installed in /etc/cooling-failure-protection/user.sh.

Enable the Beep Module for early Warnings

Load the Kernel Module:

sudo modprobe pcspkr

Then perform a Test with:

beep -f 2500 -l 2000 -r 5 -D 1000

Set the Kernel Module to be automatically loaded at Startup:

echo "pcspkr" > /etc/modules-load.d/beep.conf

Notes / Useful Commands

List Block Devices

List Block Devices (Few Properties):

lsblk -d -o TRAN,TYPE,UUID,WWN,ROTA,PATH,NAME,MODEL,KNAME,LABEL,HOTPLUG,GROUP --json | jq -r

List Block Devices (More Properties):

lsblk -d -o TRAN,TYPE,UUID,WWN,ROTA,PATH,NAME,MODEL,KNAME,LABEL,HOTPLUG,GROUP,DISK-SEQ,HCTL,MQ,PARTLABEL,PKNAME,PTUUID,REV,RM,SCHED,SUBSYSTEMS,UUID,VENDOR,ZONED --json | jq -r

Walking the /sys/devices Tree

General Principle:

grep -ri <property> /sys/devices/

For instance in the case on a SATA SSD

grep -ri <property> /sys/devices/pci0000:00/0000:00:1f.2/ata1/host0/target0:0:0/0:0:0:0/

References

• https://www.truenas.com/community/threads/script-drive-temps.90892/