Re: Thermal Monitor & automatic shutdown on over temperature event
Hi all, I have similar worries since I run a fanless EPIA. If load
gets to high and I'm not at home, or asleep I want the system to shut
down.
I wrote this small script using sensors to detect and warn high temps.
At very high temps execute shutdown.
A wall-message and a 30 sek timeout is good if I'm at the desk and can
connect a fan and abort the shutdown. I run the script with nohup.
regards, David.
#!/bin/sh
HIGH=70
SHUTDOWN_TEMP=85
export HIGH
export SHUTDOWN_TEMP
while true
do
# read temp
TEMP=`sensors|grep 'CPU Temp'|cut -d+ -f2|cut -d. -f1`
if [ "$TEMP" -gt "$HIGH" ];
then
echo "High temp: $TEMP > $HIGH" | wall
sleep 2
else
if [ "$TEMP" -gt "$SHUTDOWN_TEMP" ];
then
echo "To hot, shutting down in 30 sek.!" | wall
sleep 30
sudo shutdown now
else
echo temp OK: $TEMP
sleep 5
fi
fi
done
On Mar 9, 9:10 am, Bob <s...@homeurl.co.uk> wrote:
> Hi, I'm trying to get my etch MythTV Backend to shutdown automatically
> if either the CPU or the Hard drives overheat.
>
> My motherboard is an ECS K7S5A which has a it87-isa-0290 sensor chip
>
> from sensors-detect
> # Probing for Super-I/O at 0x2e/0x2f
> # Trying family `ITE'... Yes
> # Found `ITE IT8705F Super IO Sensors' Success!
> # (address 0x290, driver `it87')
> #
> # Probing for Super-I/O at 0x4e/0x4f
> # Trying family `ITE'... Yes
> # Found `ITE IT8705F Super IO Sensors' Success!
> # (address 0x290, driver `it87')
>
> Sadly with this motherboard it's not possible to read the on die
> thermistor without breaking out the soldering irons and modifying it.
>
> I can't find a way to set the "Overtemperature Shutdown limit" mentioned
> in the it87 documentshttp://www.gelato.unsw.edu.au/lxr/source/Documentation/hwmon/it87
> and I've tried setting temp3_over (CPU) to 40 and used burnK7 up to 53
> degreesC (actual is probably +10 ~ 15) and the system doesn't shutdown.
>
> I've seen people recommend using sensord to shutdown on overheat but
> can't see howto, also this wouldn't do anything for the hard drives,
> similarly hddtemp seems to just do reporting with no option to shutdown.
>
> collectd has options to monitor both but no tools that I can see to do
> anything if an error condition arises.
>
> My reading has led me to believe that the kernel can do at least the CPU
> part of this for you, if you've got a better supported sensor chip,
> which I don't, but I can't be alone, so what do others do in this situation?
>
> Thanks, I'm not subscribed at the moment so could you please CC me on
> any replys.
>
> --
> To UNSUBSCRIBE, email to debian-user-REQU...@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Reply to: