I am in the same boat and I am extra frustrated because I have a second HASS install at another location that I recently migrated to a different machine with zero issues. Now I have tried restoring a backup of my home setup to two different Raspberry Pis with no luck
had the same problem. A CLI restore eventually worked a bit. Most was restored. Had to reinstall a few addos, reinstall mqtt , mariadb, reconfigre zigbeed2mqtt. Finally after a full day, my ha is working.
I need help if anyone knows how to get out of this mess.
Following the last update several devices started to malfunction. So I decided to downgrade.
First I tried to restore from HA without success. As it didn’t work, I tested two methods explained in the french forum: make a clean installation (therefore erase everything) then restore from the welcome message or create an account before importing a backup and restoring.
Nothing works.
In both cases after clicking on restore, nothing more happens. I waited a whole night before restarting. Nothing changed. Therefore a new and virgin HA.
I’m on a Raspberry Pi4
HassOS
OCZ Vertex 60GB SSD
I made an disk image of my SSD under Windows, sector by sector before erasing in the first place. But when I restore this disk image, the system does not boot.
I tried :
- Complete restore on a new install
- Restore first the core then the addons and the files on a new install
- Restore the files and the addons then the core on a new install
- Clean install on SD card and restore in this system
→ In all cases only the restore of addons works
- Restore disk image of my SSD made with HDClone X.4
- Reinstall on SD card, after core update
→ The system does not start
Archives files are clean : I can open and browse it. I’m trying to restore manually by copying files, but it doesn’t work for now.
Another hassle: I can no longer connect in SSH on the RPI using Putty. I put a ‘ssh’ file at the root of the SD card, but the connection is refused (I can ping via windows). I don’t know how to access the bootloader to target the SD (and not the SSD) because it is accessible in SSH! Over the internet there is a lot of resources to switch from SD to SSD but not the other way around…
I’m considering redoing everything from scratch but before that I need a working backup system, no way to redo everything and lose everything again.
If anyone knows what to do, thanks for the help, this is my worst experience with HA since I began last february.
I’m in disbelief, but after 6 hours my restore actually completed. A 250mb backup took 6 hours. Give your restore more time and see what happens
1 Like
@ifeign Thank you so much for this post. I was hopeless, when after 2-3 hours I just started over, because couldn’t accept the fact that it might take sooooo long. Thanks to you I found patience and after 4 hours (150Mb backup) it finished successfully. For me the restore process is crap - no heartbeat, no indication of progress, nothing.
3 Likes
Hi, same problem here.
Good that I found this post and had patience to wait…left it during the night and in the next morning the backup was restored and HA running fully functional.
Before reading this post I had tried to restore several times from fresh HA installation in the believe that nothing worked…
Is there any way to track the restore process? Logs?
1 Like
My HA crashed big time yesterday when I tried to update to 2023.9.3. HA never woke up anymore after I started that update. I tried to restore from backup, but nothing happened. I found this thread and understood that it should take time (my backup is 1.9 GB), but things went a bit uncomfortable because it was getting dark and I couldn’t control my lights etc!
Anyway, I let the restore process running for full 12 hours, and still nothing. So, I started from scratch using this guide: Home Assistant: Ultimate Restore Guide thinking that since my backup was over 1GB, I could just drag-and-drop my backup file straight from Google Drive to HA (although it seemed to work at least since restoring process started). Anyway, now I tried that over 1GB size method, and although it has been running only 25 minutes now, I fear this is not working either. I think next I will try partial restore, if that works one part at time… I just have to get this working today, and if backups don’t work, then my home automation is in unacceptable state. I remember restoring from backup maybe 2 years ago, and things worked just fine, but then I had a lot smaller system. Still, it should be crucial to show user SOME PROGRESS about restoring process!
same issues occurs to me.
I am about to reinstall my supervised installation based on debian. But before I’ll do it, I started a test if restoring a backup acutally works.
Therefore I created a supervised HA VM and on the onboarding screen I did “restore from backup”. Based on CPU utilization it seems that something happens for around 30min before the system idles. The process is called "python3 -m supervisor.
But the system never gets done and after a reboot (after 24h of waiting) I get to the onboarding page again.
I also tried to create a dummy account first and then restore a backup from the settings menu. This somehow crashes the entire system and even after reboot it is not accessible anymore.
My backup file is ~1.4GB. The HA version of the system where the backup was taken and the target restore system are both 2023.9.3.
Is there any other way to restore a backup? Maybe copy all files manually etc.
Hello Same here. My 32 Gb SD Card has no more Space. My Backup is 900 MB.
I installed a fresh System on 128 GB Card and tried several times to install the Backup.
After a few hours no response from IP:8123
After a reboot the Onboarding Page is back…
Can you get to a console? SSH? And, are you running Supervisor (not sure if you can on a Pi or not)?
Go to the console and type:
supervisor logs
Mine took a while. I am running it on an old quad core server and the backups are on a Samba Share. So, it took a while. But, I expect it to take much longer on a Pi.
Here is my output at the end:
Hi.
I have a very strange problem. Let me explain it:
I’ve just made a backup from my RPI4-HA 2023.10.3 using Settings/backups in order to use it in a new instalation as a container too in another equipment.
II expected that after installing HA on the new device, there would be a “restore” option, but I only get the screen to install from scratch.
I was expecting the option that should appear, right?
I have the same problem but my backup it 3.1GB (not sure why until the latest release it was about 1.1GB). Now I struggle to backup on the Pi so I’m trying to move to a VM on a bigger system, but the backup cannot be uploaded to the new server (it errors out at 3.05 GB) or just gives a http 400 error.
Really disappointed that the backup and restore functions are so unreliable given our reliance on the tool.
One thing I was stumped with, is that a fresh install of home assistant is on
http://homassistant.local:8123 and it seemed restore of the backup did not work
The issue here was that my backed up config was from a system using
https://homassistant.local:8123 so it never came back finishing the backup
So change http to https after nothing seems to be happening anymore (wait approx an hour) and you may find your system is restored!
Just my two cents - suffering from the same issue - connected a monitor + keyboard - typing:
“core logs” - enter
(command prompt)
I see the logs are indeed changing, as if progress is made.
I had the same concerns today when I was restoring the backup. Not because it took so long but because during the process the onboarding page became unavailable. Was thinking something went wrong during restore but checking in console supervisor logs showed it’s actually doing something. Anyway my backup is 1.4GB and it took around 30-40 minutes to restore. Seems like old laptop (Lenovo T420i) with small SSD drive is thousand times faster than all this Raspberry devices. Not mentioning a lot cheaper 
I took mine actually free from my company but even if you want to buy such it would be around $50.
This statement is certainly not correct!
Raspi 5 with NVMe is not 1000 times slower, rather faster than your system
HA in Docker, importing a backup takes less than 2-3 minutes
Unpack the backup and copy it to the right place, done
It’s not only the host’s or ssd speed.
The time to restore a backup depends also on the amount of addons and the available internet download speed.
The backup contains only the addons-data, the addons themselves are downloaded from the internet.
At least if you restore the backup directly on a fresh installation.
Honestly I did restore today because I was switching from HDD to SSD. And even on HDD this restore took same time. But on SSD long period graphs (one year for example) generate much faster.
Wouldn’t it be better to clone the HDD to the SSD without importing the HA backup?
Wasn’t possible because HDD was 256GB and SSD 128GB and no tool was able to shrink partition.
Also it was just cleaner installation.