A bash script to gather system metrics like temperatures and ARC utilization that are missing from the Proxmox UI.
- The install script adds 3 packages and expects you to be on Proxmox, a Debian-based system. It adds
screenfetch
lm-sensors
andhddtemp
. If you are okay with this, run./installpkgs.sh
and follow the instructions printed at the end of the script - Copy
config.sample.sh
toconfig.sh
- Edit
config.sh
and replace the example variables with your system's disk and CPU names
3.1 Configure Disks - For ZFS users, you should be able to runsudo zfs status
and see the name for each disk in your zpool that refers to its/dev/disk/by-id/
path. You can leave the disk arrays empty or omit them entirely if you don't have mechanical drives or Optane drives, for example
3.2 Configure CPU - IF you installedlm-sensors
in step 1, you'll be able to runsensors
and find the name of your CPU. My Ryzen 3700X is shown in a comment in the sample config ask10temp-pci-00c3
and thecpu_temp_awk_print_fmt
is a format string which picks the second word in the example. It runs the format string onthe line with the matchingcpu_temp_field_label
,Tdie
for the temperature of the CPU die in my case.
3.3 Configure Screenfetch - If you did not installscreenfetch
or do not want to see it at the top of the status output, setshow_screenfetch
to false - There are some aspects of the status which are implicitly configured.
4.1 APC UPS Users - If you are on an APC brand UPS and haveapcupsd
installed with a USB serial connection to the UPS, it can read the status of you battery and show current draw in Watts and as a percentage of its VA rating. It will also show time remaining in the event of a power outage.
4.2 ZFS Users -arcstats.sh
does a lot of heavy lifting to determine if you have anarcstats
file on your system, and in turn if you have a SLOG or L2ARC and shows additional sections for them if possible. - Test by running
./status.sh
and checking that the output shows all of your disks, CPU temp, and ARC stats correctly