I see. A small powered USB hub shouldn’t be a problem.
But I won’t get a NUC . I’m just in the process of changing everything but the desktops at home from x86_64 to Arm64. Power efficiency and stuff, plus I rather have many small machines running than one big. Also, most ARM doodads can run on PoE which means, my switch can power them and I avoid cumulative losses through individual PSU inefficiency on many PSUs.
When I only had one Xeon running pretty much everything, maintenance windows weren’t very popular with the family because pretty much everything went out when I shut this thing down to do anything.
You’re shooting yourself in the foot on the NUC. Pi 5 power perf curve isn’t there. Admirable goal but…
Sure, Nuc isn’t as ‘sippy’ as a pi sure but when something isn’t fit for task it’s not fit for task. I literally don’t notice the delta on my power bill yes is it 2x the pi technically, sure but ask m how much work I ever do with it hw wise. (answer exactly once. It failed hard. I bought another and restored to it while the other was on RMA.)
(im counting the HOURS you’ve spent here and could have bought three nuc and powered them for what you would have got paid instead of working onthe issue… Just sayin’)
But I’m still eyeing that power rail and until you move those things off to self powered USB I DON’T trust your setup.
Maybe a single NUC vs. a single Pi doesn’t make a big difference but I have quite a few things running 24/7, currently spread over 7 SBCs and one low power Xeon. HA is actually the only Pi4 “in production”. Other stuff mostly runs on RK3588s which is pretty “sippy” for its power and it has an NPU which is very nice for stuff like Immich and Nextcloud. Just my infra used to bring about 600-700kWh per month to the power bill, something I very much liked to reduce because electricity is not cheap in Finland
If you follow this guide you will be able to SSH into the host system directly. From there you can simply run journalctl as normal like on a normal non-alpine linux system.
Bottom line. If you want to perform OS functions in an emergency, enable SSH natively in the OS, not a docker container Add-on. Then you won’t have these types of issues.
Average users only need plain-text logs(even if they have had to interpret, then it’s still helpful) . edit: and one can trigger actions etc upon the “local” logs
Rotating them to journal, compress them alot, and they can also be helpful for Devs ( from those which choose to share diagnostics,to Nabu )
Honestly i don’t know , thats only what comes to my mind
HA again restarted around 03:56… This time I had logs but I see nothing in them… Normal noise from the addons and then suddenly puff - the HA start log entries come up… Very odd.
I guess, I’ll start disabling addons and see how that goes…
. I’m just in the process of changing everything but the desktops at home from x86_64 to Arm64. Power efficiency and stuff, plus I rather have many small machines running than one big. Also, most ARM doodads can run on PoE which means, my switch can power them and I avoid cumulative losses through individual PSU inefficiency on many PSUs.
, thats only what comes to my mind