Almost nothing. Recently moved the core from Synology and decided to start afresh. A few local albums, Tidal account with a few or no albums in favorites.
Description of Issue
Today I was suddenly greeted with this message “There was an issue loading your database”.
Why is it a dead-end error? I have a database backup. I can restore it but there is no option for this.
I searched this forum. I tried two solutions suggested here. I tried deleting the tidal_account file in the database/registry/core folder. I tried reinstalling the Roon server. None helped. Many reported solutions were via PMs.
Any help?
P.S. Tried uninstalling the RoonServer, rebooting and re-installing. Didn’t help. Many search results point to an old issue from 2016-2017 that was supposedly fixed then.
This deserves a second post. I decided to delete the Database folder. When I started Roon it could not connect to the core, so I chose “a different” core, connected to the same core and chose restore backup. As I said the database had almost nothing in it. So it quickly restored. I almost felt happy but it never finished got stuck in this “Restoring backup Installing” screen:
I wonder what’s going on. I’ve never had so many troubles with Roon.
P.S. It’s been like this for three minutes. I am going to stop it.
P.P.S. LOL! Now it’s stuck in “canceling” screen:
I am going to try to cancel “…Canceling…”
Pressing Cancel didn’t do anything, so I closed the Roon client.
A note: during this time of “restoring” and “canceling” I observed zero CPU activity from Roon on the core PC.
I think the saga is over. After I closed the client. I closed and restarted the core. I opened the client again (I use it on a different PC). I was greeted with an option to choose the core, I chose the same core, opted to restore backup and restored a slightly earlier backup. This time it restored and prompted to relaunch. I did and it seems everything is back to normal.
It looks like it was an issue of unrecoverable database corruption. If support can confirm that this error means just that it will be easier for other users to troubleshoot in the future. I found many posts about this issue but the resolution was often inconclusive and sometimes threads ended without resolutions.
Why not add an option to restore backup on that screen?
It’s hard to know for sure without seeing logs from before the restore, but it definitely sounds like it might have been corruption based on your description.
I’d recommend running tests on the drive where Roon is stored, as well as your RAM, just to make sure there are no faults there — We’ve found that failing hardware can lead to corruption like this.
I want to post an update and mark it as a solution. I hope this thread will be un-unlisted and be helpful to others.
Several days before this happened there had been an extended power failure for several hours. The PC with the Roon server was on a UPS which obviously exhausted the battery. It was supposed to hibernate when the battery got low but the recent Windows 10 upgrade replaced Windows generic drivers with the APC drivers that disabled Windows builtin power management and the computer was just disconnected from power unexpectedly.
When the power returned the Roon database was corrupted but… Roon was running and… backing up its corrupted database. The first backup restore didn’t work because its timestamp was days after the power failure. The second backup restore from before the power failure was successful.
My recommendation to users who run into this problem: Restore the last working backup. Delete the database folder, choose connect to a different core, select the same core and choose “restore backup”.
Roon was running and backing up its corrupted database [EDIT: database, not core] on schedule. If it does it enough times without users noticing it, users may be left without working backup. Consider implementing background periodic checks of the database for integrity. Also, consider changing the format of the database, so it doesn’t get corrupted on power failure. When the power failure happened nobody was using Roon. I run other software on the same PC that constantly writes to hard drives, and the Roon database was the only thing that got corrupted.
The error I got (in my screenshot in the OP) didn’t give any option besides “let us know”. Consider giving an option to restore backup on the same screen. From what I gathered that’s pretty much the only way out of this error. If people backup regularly it will be harmless and will reduce load on support.
