I’m migrating from a Pi4 to a Pi5. I was able to backup my data very quickly, but the restore to the new device is taking forever.
Is it possible to determine why it’s taking so long and potentially speed it up? My backup file is only 160meg.
I also noticed that it took forever just to install HAOS onto it. It took over an hour for the install/initial boot to complete (they part where the screen pops up and says it’ll take over 20 minutes). I’m curious if that is related.
My specs are:
- Pi5 with 16gig of RAM
- 256gig Sandisk U3 SD Card
- PoE HAT
- Samsung 960 EVO NVMe SSD
Yes it’s over kill… it just happened to be the specs of a spare Pi5 I had
In your new device, at what moment you do the restore? Before or after the first setup steps?
Its seen more that doing initial boarding , makes a quicker restore
I clicked the restore button as the very first step. I think my options were something like ‘setup your home assistant’ or ‘restore’. I choose restore, then selected the backup file, and have now been waiting for hours and hours.
I wish there was a log or some type of status to monitor. Either there’s something wrong with my device or this is the most horrible segment of code ever written.
Have you refreshed your browser? That’s a super-common issue where it finishes but the page doesn’t get automatically updated to reflect this.
Try again but then after first boarding, it Will be done quickly
If I had to place money down, I would say it’s something on my side. I’ll check back once I’m by my machine next. There’s no way it can be this bad normally.
I forgot to close the loop on this. I think the issue is that the restore process was hitting my home assistant at “http://homeassistant:8123”. That URL wasn’t valid once the restore was completed (I use a different port and such).
I suspect the script failed to notify me that it was done, and a refresh wouldn’t have worked, because the ‘restore URL’ wasn’t the same as my ‘production URL’.
In hindsight, if http://homeassistant:8123 was valid post-restore, things may have worked out better for me. I haven’t had a chance to blow away my setup and re-test a restore. I hope to in the future.
Alternatively, if the restore script was ‘smart’ enough to look at the configured ‘internal URL’ and automatically redirect to it once the restore was completed… that may have worked as well.