inspired by zpscan.sh and fixer.sh, I wrote myself a script to monitor disk failure and blink LED using
How to use:
diskfailLED.sh
Scan a pool, blink LEDs to locate bad/missing disks, unblink for good ones.
First try something like
Note that the script would only attempt to turn off LEDs of good drives. So the LED will keep blinking if the bad drive is removed and slot left empty,etc. In this case put the replacement drive into the same slot or use
sesutil
, which I feels much easier to use comparing to sas2ircu
. It also shows wider compatibility in my limited testing, it worked for both my Dell R620 and NetApp DS4243, which didn't work with sas2ircu
. I don't have a faulted disk at hand so I only used zpool offline
to simulate disk failure and it appears to work fine.How to use:
diskfailLED.sh
Scan a pool, blink LEDs to locate bad/missing disks, unblink for good ones.
First try something like
sesutil locate da1 on
on your system and make sure it's blinking the right LED. Then put the script into cron to run every minute or so.Note that the script would only attempt to turn off LEDs of good drives. So the LED will keep blinking if the bad drive is removed and slot left empty,etc. In this case put the replacement drive into the same slot or use
sesutil locate all off
to turn off all LEDs.