Got it. I found a javascript error using right-click -> Inspect (Q) in Firefox:
Error: $scope.configureBackup is undefined
$scope.restore@https://<IP_ADDRESS>/js/restore.js:290:13
[...]
That finally led me onto the right track, it's a bug that was fixed in v7.1 by commit 7a6a9cdb: https://git.cloudron.io/cloudron/box/-/commit/7a6a9cdbb4374eae9051326a51a4da21409d662a
I manually fixed that line in restore.js on the server and now it seems to work fine.
I'm trying to move my cloudron to a different provider, following the instructions in the Cloudron Docs. After installing Cloudron, uploading the Backup config, and clicking on the "Restore" button, I get to a page with a spinning progress indicator and three dots below ("..."). After that, nothing seems to happen at all, even if I leave it for hours. I don't see any activity in the browser window, I don't see anything happening using journalctl -f, the boxdata and appsdata folders remain empty. I don't really know where to look for any error messages.
Some info that may be relevant:
I don't really know where to investigate or what to try, especially since I can't find any error messages to give me any hint of what's wrong.
Could the mismatch of Ubuntu versions be a problem? I was hoping that as long as the Cloudron version matches, it should be fine.
Is there any place where I might find logs or error messages related to restoring a backup?
Is there any chance that browser settings have any influence (cookies, ad blockers, ...)?
Do I have to stop the original Cloudron?
@girish Thanks, that works.
I use rsync backups on a local file system (details below). I have just noticed that my backups consume a lot less space that I'd expect - 1.2 GB for all backups combined, when my Nextcloud app alone has over 6 GB of data in its "appsdata" folder. It seems that Cloudron only backs up the files that have changed since the last backup, without ever checking whether that backup still exists.
How can I force Cloudron to do a full backup, including files that were previously backed up and have not changed since?
Some details on my backup strategy for context: my final backup location is a sever I have at home that I don't want to expose to the internet. So I tell Cloudron to do backups to its local file system, and then I have rsnapshot on my home server fetch those backups. Occasionally when I run low on storage on the Cloudron server, I delete old backups there. This worked fine for years; I assume a (fairly) recent Cloudron update introduced the change that Cloudron explicitly excludes unchanged files from backups. (Which seems rather pointless when using rsync on a local drive, because rsync can take care of that itself, but I suppose it makes sense for many other backup cofigurations).
I had some trouble with my Cloudron. Found out that the disk was getting full, and managed to fix that with the instructions at https://cloudron.io/documentation/server/#recovery-after-disk-full
However, I investigated a bit what was taking up so much space. It turns out that by far the biggest culprit is /var/lib/docker/overlay2 which has about 450 subfolders and uses 16 GB of my 45 GB drive. (determined with "du -x", i.e. skipping anything that's not on the root file system).
I use only 6 apps, and the appsdata folder is only about 7.5 GB. Is it normal that the /var/lib/docker/overlay2 takes up twice as much space as my appsdata? My understanding is that that folder contains the docker containers of the apps and whatever other containers are running, so it seems odd that it's that much larger than appsdata.
Is it possible that some old content is not cleaned up or duplicated when the VPS is rebooted? I had to reboot a few days ago (for the first time in months), just wondering if that has any connection. It was a hard reboot, i.e. equivalent to just cutting the power without shutting down.