I’m running HA in a docker env on my Pi.
What’s the best way to do system monitoring including monitoring the docker containers?
The systemmonitorsensor platform unfortunately does not know the containers and the filesize sensor can’t watch files on the host itself (outside the /config path).
Glances knows containers, but adds 5% CPU usage just for monitoring.
(And my HA does not recognize the glance platform. In the log I see Platform glances not ready yet although I can open the glances web site)
fritzbox_netmonitor to have more details of network stuff.
Does this provide further info not included in other sensors but useful?
speedtest produces a lot of network traffic just for monitoring. Useful? Cheaper options?
No way to get filesize of files outside /config folder?
Tried with command sensor, but seems that those also do not see the files outside /config.
- platform: command_line
name: A_File_Size
command: "du -m /home/pi/example.jpg | cut -f1"
unit_of_measurement: 'MB'
value_template: '{{ value | int }}
But this command sensor does work, although it reads a file outside /config:
- platform: command_line
name: cCPU Temperature # CPU Temperatur in °C
command: "cat /sys/class/thermal/thermal_zone0/temp"
unit_of_measurement: "°C" # If errors occur, remove degree symbol
value_template: '{{ value | multiply(0.001) | round(1) }}'
You can access stuff outside of /config because docker links them by itself (system stuff often gets linked because programs use it). If you want to monitor the whole hard drive or whatever you have, you’ll need to ‘mount’/link your drive’s root directory (/) into the container and then monitor that with HA. You could mount it wherever you want, just be sure to not overwrite stuff. Try: /srv/drive or so, that should be free