it’s a kernel crash in the dlls drivers for network / bluetooth coms
That does not look like a network driver issue, to me that looks like a bug in the CPU scheduler
Just adding a data point here. As I posted earlier in this thread, my HA would stop working randomly. Yesterday it died entirely - wouldn’t boot up again. Discovered that the culprit was either the SSD (a SanDisk M.2 SSD) or the Unitek USB housing for the SSD. One or both of them are now dead.
This may have been what has been causing the troubles earlier.
I’ve replaced both with a Samsung 2.5" SSD and a SATA/USB connector, and restored a full backup from September. Everything is now back up and running, but I do have a gap of historical data from September to yesterday…
You should consider moving your database to a legit server. I’ve had a synology (cheap ~$200) server with 6TB of storage running for 7 years. It contains my database. No issues whatsoever. Not to mention, you can use the server for other things like music, movies, cctv recordings, and personal documents.
I think every HA installation should have the Home Assistant Google Drive Backup addon. Automatically takes the backup every day and uploads it to google drive. Check it out.
1 Like
That seems like a step backwards for a system designed to be independent of the internet.
You can take a flash drive copy and bring it manually to a site of a choice, but having one of your backup offsite is a basis the commonly used “3-2-1” backup strategy (3 backups, 2 medias, 1 offsite).
For sure, having a single backup on the same machine you backup from is not a great stategy.
Awesome suggestion, there is also GitHub - ikifar2012/remote-backup-addon: Automatically create and backup HA snapshots using SCP for those (like me) that cringe when there isn’t a self-hosted option.
This is happening to me at least once a week as well. I am using a Home Assistant Blue running the latest core version and 2022.7.3. This has been happening to me for 2-3 months now.
Everything feels cool to the touch. The heat sink sits on top of a fan so I don’t think there is a cooling issue:
I’m also seeing this with Home Assistant Blue for the last few months, always running the latest versions, but usually skip vX.X.0’s
Similarly, if I disconnect my ethernet physically for ~10 seconds, and reconnect, it comes back online.
I’m able to ping my host IP fine, while it’s down, but can’t access homeassistant.local.
I’m running on Nest Routers/Mesh, with HA connected via a switch. No other device on that switch or in the network has the same issue.
Strangely, mine also seems to go down only on late Wednesday’s/Thursdays (Ireland). Last outage was a few weeks ago.
Perhaps relevant. I can see the ethernet led on the HA blue flashing, and can ping no problem, but all of my automations and zigbee are down, which knocks out my lights. So that’d suggest to me there’s something going on with the container/supervisor/core.
And in terms of usage, If anything I seem to be under utilising odroid 
I’ve got my HA’s IP set to static, with a designation in Nest/Google Wifi, which has always been the case for the last year. IPv6 is disabled.
My network settings:
Do you have http entry in your config file? I don’t have one and wondering if that has anything to do with the issue.
@alexcapone I don’t have http in my configuration.yml, never have.
Since posting, I’ve had about three instances of home assistant acting strange:
- Speedtest.net sensor reports ~5mb download speeds (Usually >450mb)
- Can’t connect to homeassistant.local sometimes, but can access via Nabu Casa
- Disconnecting ethernet, restarting router/network, restarting HA core and/or HA supervisor, did not fix these issues the last time. Only a physical reboot worked (removing power cable for a few seconds)
I have the same problem with my rpi4 on ssd.
I think it has something to do with the bluetooth. I rly want to get this problem fixed because when im on hollydays the people in my house gettin rly pissed by this.
Havent found anything in the logs everything says that this and this is not working anymore but there is no initial thing that triggers this behavior.(at least i didnt saw it)
Did someone solved it already?
Still seems to be an issue. I have Raspby 4 4GB RAM and SSD via USB in a place which keeps just stopping knocking the whole system out. When I boot it up all log files start from scratch. Is there a way to not do that so I could see what latest log entry was before crash? History is also gone every time. At home I have a very similar set up (same Raspberry pi specs) and it never stops. Difference to stopping one is Zigbee stick (conbee vs sonoff) and a slightly different SSD hardware. My case has a cooler for the CPU but not for the SSD section. I have no answer to it yet 