I am running rpi4 using original 3a power supply with SDD powered off the RPI with no power problems. You just need to do the math on all the peripherals you connect to ensure your external devices do not draw more than 1.2 A
Power Supply - Raspberry Pi Documentation.
After you do the math, you may be ok, if not you can always go with a powered USB hub.
In my case, sata ssd 700mA + USB2sata adapter 100 mA + Insteon PLM 0mA + Argon One Fan 100 mA
Total = 900 mA
I was able to update (rpi4) without any issues. However, before the upgrade, my CPU was pretty steady at around 20-25% pretty much all the time. All this installation does is manage the Z-Wave and Zigbee devices via a HUSBZB-1 dongle. Since the upgrade, every 5-10 minutes the CPU use briefly spikes to 50-70%.
Which version 5.x have you been running before?
What you are seeing is the EFI shell provided by your virtualization system. It seems that the boot partition got corrupted, and your BIOS isnāt able to load the boot loader.
If you have a recent snapshot, a new installation and restore from a snapshot is probably the best way forward.
What Virtualization System are you using? There are instructions how to recover the boot partitions for some installations, e.g.
If KVM/qcow2, this guide might help: Upgrade to OS 5.9 on OVA caused corrupted EFI folder Ā· Issue #1125 Ā· home-assistant/operating-system Ā· GitHub
Same issue here, hassos on VM (synology) on SSD. I checked disk and itās fine.
Working again after downgrade:
os update --version 5.13
Does someone knows what this means:
[16:06:24] ERROR: The configuration check did not pass!
[16:06:24] ERROR: See the output below for more details.
Testing configuration at /tmp/config
Fatal error while loading config: [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
Failed config
General Errors:
- [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
What do I have to do?
I have the same error as well. Though I did a full backup and upgraded to os 6 successfully. Before updating from core core-2021.6.3 to core-2021.6.4 I see this error from the Check Home Assistant configuration (v3.7.1).
[20:00:33] INFO: Checking your configuration against this version...
[20:01:50] ERROR: The configuration check did not pass!
[20:01:50] ERROR: See the output below for more details.
Testing configuration at /tmp/config
Fatal error while loading config: [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
Failed config
General Errors:
- [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
Successful config (partial)
[20:01:50] INFO: The full output has been written to /share/check_config.txt
I then checked the /share/check_config.txt and this is what it says:
INFO:homeassistant.util.package:Attempting install of colorlog==5.0.1
[01mTesting configuration at /tmp/config[0m
INFO:homeassistant.util.package:Attempting install of mutagen==1.45.1
INFO:homeassistant.util.package:Attempting install of aiohttp_cors==0.7.0
INFO:homeassistant.util.package:Attempting install of gTTS==2.2.2
INFO:homeassistant.util.package:Attempting install of async-upnp-client==0.18.0
INFO:homeassistant.util.package:Attempting install of netdisco==2.8.3
[31mFatal error while loading config: [0m [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
[01;37mFailed config [0m
[01;31mGeneral Errors: [0m [31m [31m
- [Errno 2] No such file or directory: '/usr/local/lib/python3.8/site-packages/idna-3.2.dist-info/METADATA'
[0m
[01;37mSuccessful config (partial) [0m
Figured Iād wait to see what others say, or another version comes out, before upgrading core.
Youād lose the autofill entities of home assistant showing up in Node Red, wouldnāt you?
Not that i have noticed.
There is a homeassistant contrib
I dont know if this is the correct one. Just the first google hit. I can check later today
Update: Yep, this is the one I loaded into NR. And it just autofills.
Just updated my Rpi4 4GB running off a Sabrent external enclosure and Crucial 120GB SSD. No issues at all so far. Dropped 5 pings as it rebooted then HA started up shortly afterwards. Whole process took less than 10 minutes 
My VM is running on a QNAP NAS in the Virtualisation Station is actually QEMU.
I started to figure out that it was a problem with the VM image and not Home Assistant, so snapshots wouldnāt have saved me that much, so @agners you are spot on there.
I then switched to trying to recover configuration files (including snapshots) from the disk image itself. After trying a few recover apps, I found that DMDE was able to open the disk image and pull out files from the image - complete the full file names and the directory structure in tact, so worth the purchase price.
I was able to pull out snapshots of individual add-ons, HACs components, my NodeRED config and other stuff, installing from a new VM image of Home Assistant OS Release 6.
Everything is now pretty much back to normal.
Iām going to do more scheduled backups from now on, complete with more image backups of the VM 
Thanks @quadhammer @basesnow @nontijt and @agners for your tips 
No problem.
Last night I did exactly the same thing.
I had a slightly older backup of the HA VM. Put that one back. Created a full snapshot and copy it.
Then I downloaded a new .qcow2 image from the download page you mentioned. Installed the new image using the same MAC adres etc. start it, install samba addon, past the full snapshot in the folder back again, restore and 1+ hours later it was back up running again. 
I am using proxmox (type 1 hypervisor). HA OS in a VM. I have a separate container for mqtt, node red, zwaveJS2mqtt. So basically the HA is only the frontend, which works perfectly for the past couple of months. fast as hell.
The only question I still have, why/what is the cause of the problem in the first place. ?
Ah, computers! Iāll probably move it all off the NAS at some point and onto a NUC. Itās all a bit of amusement 
Gheheheā¦ agreedā¦ Home Assistant is my hobby. a SERIOUS hobby ofc. 
Iām using Oracle VM virtual box, which uses a different image, and Iām having the same problem, so it canāt be just a problem with the KVM virtual machine.
I was using a āheadless startā, so I couldnāt see this error in the background. Using the virtual boxmanager, right clicking on the OS and picking āshowā, brought out the window above, and then when I pressed control D to continue, it updated to OS release 6.
Not sure what those other two errors are, or if they are related, but I hope that helps someone, even if it just lets you get into the snapshots.
Edit: I ended up installing the haos_ova-6.0.vdi from scratch. It was pretty painless, the only caveat was that you had to make a fake/temporary account when first loading home assistant, before you could restore the snapshot. It was good that you didnāt need to install Samba, (thereās an option to upload your snapshot), but you still need to make an account first which is a bit stupid.
Nice to have it auto loading again on start-up.
been getting Error 404 on 120gb ssd with Raspberry PI 4 since update. Tried a fresh install and now getting it on startup, tried downgrading same thing. 
Hi, did you ever figure this out?
Iām trying to get my USB Coral accelerator to work on HaOS to support Frigate but without any successā¦
Hello! Iām one of the people who had stability problems running releases after 5.3 on a Raspberry Pi. I set up a spare Pi with a mirror installation, updated to 6.0, and it ran without issues for a day!
Seems promising to me, since it crashed every night before. I will keep it running for a week or so, and post any new findings.
not really: currently I am running frigate directly on proxmox host (mainly to minimize inference time).
It is working great so I think I will keep it like this
BTW to me it is not clear if this OS 6.0 upgrade is safe or if it is better to waitā¦
After updating OS to 6.0, network (WiFi) became extremely slow. Many timeouts in logs - for all services using network connections (even local ones). HA became simply unusable. After downgrading to 5.13, everything started working again. Then I joined beta channel and upgraded to 6.1. No luck - still many timeouts; SSH login, opening Supervisor / HACS tabs etc. takes ages (but finally succeeds). Downgraded again to 5.13, and again everything is working flawlesslyā¦ Have no idea how to diagnose, whatās going on with 6.0 in my caseā¦
Raspberry Pi 4 Model B Rev 1.4, 4GB RAM, Core 2021.6.6
Just a positive update: running HASS OS 6.0 for 10 days without any issues so far. Upgrade went smooth, no hickups afterwards. Pi 4 B 8 GB a lil bit overclocked with ethernet connected network on SD card (I bet all the SSD users suffer on HASS OS updates? CouldnĀ“t manage to migrate from SD to SSD yet unfortunately).
Update: Latest HASS OS 6.1 update went flawlessly too.
