On Valentine's Day, I had a situation where the fan on one of my hotswap cages was failing, and the drives began to over heat.  My case (Rosewill RSV-L4411) does not have fan failure alarms, nor does my mobo have enough headers to be able to monitor those fans with it.

 

Since Dynamix has some issues with notifications when not logged into the webUI, I banged together a script to notify the user via the syslog, and audible beeps from the server.  (I left email reporting to be handled via Dynamix, under the assumption that it will (or soon will) work correctly.)

 

There are a number of user-configurable variables to tune the operation to your own tastes:

 

POLLTIME=300
WARNPOLL=30
SYSBEEP=1
ALWAYSBEEP=1
LOGGING=1
ALWAYSLOG=1
LOCAL=1
ALWAYSECHO=1
NOTIFY=1
ALWAYSNTFY=1

 

POLLTIME is the amount of time in seconds between scans when the temperatures are normal

WARNPOLL is the amount of time in seconds between scans when the temperature is above the warning threshold

SYSBEEP when set to 1 beeps the system's speaker.  Set to 0 for silent.

ALWAYSBEEP set to 1 to continually beep the speaker every WARNPOLL seconds when above the threshold.  Set to 0 to only beep when a transition from normal to warning occurs

LOGGING set to 1 to log the drive and temperature to the syslog.  Set to 0 for no logging

ALWAYSLOG set to 1 to continually update the syslog every WARNPOLL seconds when above the threshold.  Set to 0 to only log when a transition from normal to warning occurs.

LOCAL set to 1 to log the drive and temperature to the local console.  Set to 0 for no logging

ALWAYSECHO set to 1 to continually update the local console every WARNPOLL seconds when above the threshold.  Set to 0 to only log when a transition from normal to warning occurs

NOTIFY set to 1 to use the notification settings in the Dynamix webGUI to send out email / browser alerts.  Set to 0 to not send out notifications.

ALWAYSNTFY set to 1 to continually send out the notifications every WARNPOLL seconds when above the threshold.  Set to 0 to only notify when a transition from normal to warning occurs.

 

The thresholds are set via Dynamix (settings/Display Settings)

 

The Warning Disk Temperature sets the threshold

If / When the drive(s) reach Critical Disk Temperature, then the script will perform a clean powerdown.

 

Because of the fluid nature of the GUI right now, I couldn't be bothered to make this into a plugin, since the next beta or two might render all of this redundant, so right now you just have to extract the files to your /boot/config folder and modify the .cfg accordingly, and add the following to go

 

/boot/config/overtemp.sh&

 

 

 

Overtemp.zip

There's a band-aid fix posted by Bonienl which handles notifications now through the GUI

 

http://lime-technology.com/forum/index.php?topic=38193.msg354039#msg354039