This is a script to monitor your ZFS pools in TrueNAS SCALE, and light up the FAULT LED on compatible SAS backplanes when a disk fails or otherwise goes offline.
This script has had very minimal testing and should not be relied on for production use.
This script is only intended to run under TrueNAS SCALE, and has only been tested under version 23.10.1. It will not work under TrueNAS CORE. It may work under other Linux installations, but most likely only if you have created your pool using gptid designators.
This script relies on encled. git clone that repository to some place on your server.
git clone this repository to a convenient place on your system.
Change to this repository's directory, and cp zpscan-config.example zpscan-config. Then edit zpscan-config and set the path to the encled script there.
In the TrueNAS web UI, set a cron job to run the zpscan-scale.sh script on your desired schedule; I'd suggest somewhere between every 15 minutes and every hour.
As noted above, this script has received minimal testing. It's been shown to work with a Supermicro SAS2 expander backplane on one system. It's expected, but hasn't been shown, to work with other SAS backplanes. It's unlikely that it will work when a disk has completely failed to the extent that it's no longer seen by the kernel.