Weirdly I have a somewhat strange issue with HA, my server that deployed on a Raspberry Pi 3 and it has 80% free space on memory storage, and everything works fine, but after a few days the CPU temperature goes up and up, till about 7 days, after that my server hang up and I have to reset it with removing the power. What’s going on with my HA?
I just have a few integrations (about 10) and not that much automation (something about 15) and all of them are default and standard things that HA officially provided. So, any idea or suggestion?
Try looking in the logs. When Home Assistant starts, it renames the old log so in the config directory you will find a log file: /config/home-assistant.log.1 see what errors are shown in the log from right before it stopped responding.
It’s not something that happens in a moment, it starts exactly right after the start-up and I think takes up CPU (that causes high CPU usage after a few days), and it continuously raises CPU usage. Seems a process opened and never closed, and another one opened after that and never closed either, and so on.
Logs don’t persist a reboot and you can’t stream them externally via syslog (this needs t be fixed…).
I had I diagnose by plugging into a monitor and watching the console. I work from home so just watched a secondary monitor until the console started spewing text and took a picture before it hung.
Formatting didn’t fix and I was on a SSD so didn’t suspect a bad disk.
Ultimately it was memory read errors I think. Either way I replaced with a pi4 with more memory and didn’t hav any issues.
May just want to see how a pi4 fares and save the headache.
I just moved my setup from rpi3b+ to an extra lenovo b50 laptop, because of instability.
Until I got the transfer done, I ran node-red script that pinged my LAN router once in 5 minutes, and if there was no response, the script rebooted the rpi.Ugly as heck and my lights misbehaved after reboots but the system always came back up after reboot.
