I have an issue with my “Core” installation of HA: it keeps freezing from time to time (sometimes few times a day, sometimes a month without).
There is no error in the logfile, actually the logfile writing stops (normally, every minute, fritzhome integration logs some status in the logfile).
There are no corresponding error messages in syslog nor in dmesg.
Thiy happens with production and beta version, as already said, some quite some time by now. Drives my wife mad (she loves it when light switches etc. don’t work…).
Any suggestions?
My environment:
Raspi 4, 4 GB RAM.
Ubuntu 22.04.5 LTS
Core installation in venv with python 3.12, 2025.1.0
@jackjourneyman Not that I would be aware of. System log and dmesg do not show the usual memory low signs (cache flushes, process killing etc - nothing of that kind happens), system load is around 0.75-1.3.
@stevemann Web frontend inaccessible, mobile app inaccessible. Addons which usually regularly log to the logfile, stop logging. Node-red and regular automations don’t trigger.
Linux System itself remains accessible and healthy, restarting the service fixes the issue (until it repeats).
@NathanCu nothing unusual. Just as if everting stopped happening, log entries just stop to appear.
System log (/var/log/syslog, dmesg) state nothing unusual.
I can ping it, I can ssh to it, access other services (nodered, zigbee2mqtt) etc.
Web frontend of HA just times out, same as mobile application.
Last entry in log.1 is pyfritzhome updating devices - which it does every minute, actually even more frequently.
Yes, when HA (only HA, not the server itself) freezes or hangs or becomes unresponsive, whatever you prefer, I can still ping and ssh it.
Yes, tried a different browser, tried clearing the cache, tried everything.
It does not look anywhere near a frontend issue (because backend addons like pyfritzhome stops logging and also pressing Zigbee light switches is not propagated to Nodered anymore etc., all of this does not involve any frontend), but neither underlying server issue as well (because ssh works, other services work, no messages at that time in /var/log/syslog nor in dmesg).
Dear all,
It seems that I have the same problem. Were you able to solve it?
I am running home assistant in docker, and all other dockers etc are still running. For me, updating to 2025.1, home assistant just freezes in browser / mobile app. Similarly, there are no logs indicating a problem. I am now monitoring the CPU of the home assistant docker.
I have had the same issue for years. Have tried rebooting every night. Tried the ping integration, where it should reboot itself if it didn’t get a response. Nothing has helped. And as you state, nothing in logs…. Also on a Raspberry Pi 4. I have several times considered other hardware until I see posts with people experiencing the same thing - but on their NUCs… So yearh, I am also getting a bit desperate.
No idea if this’ll help anyone, my particular issue was fixed by updating my OS and Docker, was running Ubuntu 23.10 with the latest Docker for that OS (27.1.2?), now upgraded to the latest versions (OS and Docker) and no longer having the freezes.
I have the same issue. The problem is it runs for even few days without problem, then CPU load spikes and everything freezes. No automations, no apps, no web UI (timeout) just 100% cpu load.
I tried strace, but despite some futex operations I see nothing significant there either.
I have a dockerized setup with core 2025.1.2 / frontend 20250109.0.
I will try upgrading to current version and see whether this helps. If not, then I will try to collect some more infos here.
Same random (several times/day) issue as everyone is describing. I am running HA on an RPi5 with an NVME SSD using the standard RPI5 27w power supply. Doing a hard (power off/on) corrects the issue only to reoccur. Nothing in logs.
