Guys anyone has an idea why I can’t update core? I’m stuck on 0.118.4, I’ve tried getting help from all platforms.
I posted this in another thread, but thought maybe better answer here.
I have RPi 4, I have OS 5.3 (beta) installed. Everything is running from a tested and working rather nice SSD drive with NO SD card. Everything is VERY stable at this point.
I see 5.10 release is now available. That being said, I read some of the other comments about 5.8/5.9 upgrade on this type of setup that SSD will no longer boot, etc and requires linux knowledge or reinstalls to fix.
I do NOT possess that indepth knowledge of linux to get me out of a bind and I’d rather Not have to rebuild, restore backups, etc.
Is 5.10 safe to upgrade to and still have everything boot back up and be happy or should I just stay on the 5.3 I currently have and not do a darn thing?
This is very hard to answer. For many people there are no issues with both the 5.10 and now 5.11 updates. Others like myself have system freezes on any upgrades above 5.4. There is a github issue 1119 that has over 200 comments. If you have an issue after updating it is fairly easy to go back. Make sure you take a snapshot prior to the update and if it does not work, restore the snapshot. My issue is just the OS and after I try the update and if it freezes my Pi4, I do a restore of the OS5.4 from an SSH shell via command line.
I second Bill’s assessment. But stay away from releases 5.6 to 5.8. These 3 versions will change your boot EEPROM code, so you cannot simply roll back.
so I could do this:
ha os update --version 5.10
and if there are issues, then do
ha os update --version 5.3
I do have backups daily and I have the Terminal/SSH installed.
and it should put me right back where I was?
I have done this multiple times with zero issues.
I think you may want to add on your guide dtparam=sd_poll_once=on and how to do this. Also a mention of the quirks for controller issues. I was using recommended hardware and I was having errors (that I didn’t know of) that this corrected. The bugs with the SSD / controllers are not fully resolved.
Makes the install more complicated, but hopefully more reliable. I am working my freeze issue and when / if it gets resolved, I’ll post here so you can capture it for others.
well. This didn’t go well… She no longer boots. 
My SSD with raspbian boots fine.
Guess I’ll be reflashing the works, restoring from backup.
How long did you wait? I always leave the ssh window open until it closes. I then keep trying to open a new window while it is rebooting to watch what is happening, ha core logs, su logs, etc
Over an hour then I plugged in a monitor, keyboard mouse and nothing.
Did you unplug it? Should see something on the monitor when power is restored. Never had this issue on a working system. I had this when I had SD problems or not enough space on the SD. Normally when you install via cli, you see a response, i.e this was successful. And then several minutes go by and the system reboots. The new versions of Home Assistant are very good at repair themselves. If they have problems coming up they will restore to the previous version.
Yup, after the hour i did, slapped on the raspian ssd and it booted right up. Shutdown, put back the HA ssd, and nothing.
Well, that’s why I have back-ups right?
What SSD, What controller? If you read the github 1119 issue
https://github.com/home-assistant/operating-system/issues/1119
You will see the different issues people have had. There is a spreadsheet showing many of the different configurations that are having issues. Updates to the OS after 5.4 changed the Kernel. I have a system freeze on any of them with the new kernel. Others have different issues. I believe the boot issue you have is a controller issue and it is handled by a “quirk” setting. I know it shouldn’t be this hard, but it is in some cases. A lot of smart people here that can help get you back up and running.
Sorry to hear that it didn’t go well.
Fingers crossed that you get back to a working system soon.
On what boot EEPROM version are you? Tried the latest stable?
Restore worked nicely. On a jan firmware, was on sept. No clue what caused the issue, but it is up and happy now with 5.10, ssd, latest firmware. Whomever wrote the backup to google drive item, i owe them a beer!
So. 1 more question. I have a spare argon one m.2 case with a 256gb ssd looking for a home. Anyone try this combo for HA. Would be cleaner looking than the startech cable and ssd drive hanging out.
So 5.10 works with external SSD? Wanting to do this, Etcher gave me like 8 partition and nowhere close to the drive capacity.
Hi everyone,
Firstly thanks to @Jpsy for this great guide!
I successfully installed HA with a X857 mSATA shield running on a RPI4 4gb.
However, when I restore my snapshot from my older system (which was always up-to-date), the snapshot failes partially: the menu’s “supervisor”, “configurator” and “debugpanel” disappear making it unusable.
I’ve tried it three times without success.
The snapshots are always loaded after creating an account, as mentioned in the guide.
Is there something I’m missing?
Thanks!
I had a KingSpec 120GB SSD running for a full three months, before it died on me.
Biggest crap made east of the Chinese border.
I am using a RPi 4 running HA in Docker based on a IOTstack configuration. I always kept at least 100GB free for wear leveling, but it obviously did not help. That is, if wear leveling is implemented at all.
EDIT: if some experience a longer lifespan, then it points to sh*tty quality control, and it’s luck of the draw whether yours will last or not. Spend the dollar and get yourself a proper SSD from a reputable brand.
My experience differs to this, I have no issue whatsoever and have since purchased another USB-C device as well as a M.2 2242 drive for a thin client. All working well.
Buy another 128gb SSD. They’re cheap as chips nowadays and then install 5.11 on that new one. See how far you’ll get.