Since a couple of weeks HA (32bit on Raspberry Pi3 B+) stops working - first partially (sensors stop to report values - or more likely HA doesn’t record them) and then I can’t even access it through the browser or app.
Usually it starts by automations (or NodeRed flows) taking longer than usual to react - for example simply turning lights on or off.
What could be the problem? where could I look for errors? in the supervisor > system > supervisor logs everything is green - also the logs seem to only go back to the system boot…
I’ll upgrade to a RPi4 soon anyway and will completely set HA up from scratch (right now I’ve always just updated since 0.75 or something) - but until then, it has to work! it’s not fun comming home to a house thats almost 40°C, because HA stopped working and left the heaters on for a couple of hours
EDIT: it runs from a completely new (about 2 weeks) SANDISK Extreme microSDXC UHS-I 64 GB A2 SD card and the recorder is set up to store it’s stuff on a NAS… so I doubt it’s a failing SD card
Have a look at the memory and cpu usage in the supervisor system panel when things start slowing down. This does not auto-refresh so you have to either navigate away and back or refresh the page for an update.
A better idea would be to install some system monitor sensors and keep an eye on them (or get them to send you notifications using numeric state triggered automations).
Usually I only realize something is going on, when I come home to a freezing or hot house and can’t log in any more - and then I remember that light switches did feel slow lately. So I wans’t able to look at memory or cpu usage in that situation. I have to pull the plug and then everything works as if nothing has happened.
I’ll do that - didn’t realize this existed.
Anywhere else I can find system logs from the time the sensors stoped working?