Reboot/Restart Via MQTT

[MagicMountainMan]

Is your feature request related to a problem? Please describe.
Sometimes I need to reboot the device or restart the room-assistant service because room-assistant stops updating Home Assistant and usually stops detecting BLE devices.

Describe the solution you'd like
I would love a quick and easy way to restart the room-assistant service from Home Assistant. Ex: Send a command to restart the room-assistant service via MQTT and once the service successfully restarts it broadcasts via MQTT that it has restarted. If Home Assistant never sees that room-assistant gets successfully restarted then I could set up a notification to let me know and manually intervene.

Describe alternatives you've considered
I'm creating an automated way to restart the service with Node-Red and SSH but I haven't gotten it working yet and there would be no actual feedback to know if the service restart was successful or not anyway.

Overall I've found that occasionally restarting the room-assistant service on my RPiZeroWs makes then much more stable. I do have to completely reboot the devices sometimes, but that's beyond the scope of what a room-assistant enhancement could achieve anyway.

[saif-p]

I took your comment about using node-red / ssh and came up with a solution that works well enough for me:

• In home assistant I created an input_boolean.reboot_room_assistant_nodes and an input_select.room_assistant_nodes which contains options for 'yellowpi', ;bluepi', 'redpi' , 'greenpi', and 'all' (the hostnames for each of my pi0ws)

• In node red I have a flow that triggers when the input_boolean is turned on that checks the state of the input_select, sets the payload to 'sudo reboot' and sends that to the ssh node for the corresponding pi. It also turns the input_boolean back to off.

For the 'all' input_select, the flow sends the 'sudo reboot' payload to all 4 of the ssh nodes, with a 5 second delay between each one. I set the delay because I found that rebooting them all at once sometimes caused them to form into seperate clusters (I'm using the default cluster config).

Back in home assistant, I have an entities card that shows the cluster leader for each pi, so that as each pi gets rebooted, the state changes to unavailable.

[worm5406]

OK. I am curious. Explain more, please.

What does it reference? Does it check timestamp? What does the ssh look/send? My Pi's lag at times for 30 seconds on login etc. When it is dead/dying.

Sorry for just a few questions. I have more.

[worm5406]

I get an error also:

Permission denied, please try again.
Permission denied, please try again.
Redacted@192.168.30.2: Permission denied (publickey,password).