Roon has been running for years very well within a docker container on my Synology NAS. From sometime in December 2023 I have started to get corrupt database every so often but a backup has always recovered the situation.
Two weeks ago (approximately) a Roon update came through and I accepted this as usual and the update process advised that the database needed an upgrade. However, this ended with a message that the library couldn't be loaded and the "Help Center was there to Help" (or some such). Following directions I recovered Roon by moving the RoonServer directory and having it rebuild. This at least got me back to the screen where I could restore a back. But the last 5 backups that I have tried are apparently all corrupt, or at least completely disable Roon requiring the RoonServer Directory treatment.
I am rebuilding the database right now but I have lost all playlists and other settings.
I have two questions: 1. Was there a problem with the install or an incompatibility between old backups and the new version to the extend that old backups not upgraded as a part of the install process are invalid (now)? 2. Is there any way of restoring the backup as I have tried everything I know how to do and found no articles (quickly) that note the same issue?
Describe your network setup
Synology NAS DSM 7 using Steef de Bruijn's container image. Database is mirrored on two 1TB SSDs.
Come to think of it, at least for the more frequent database failures I have experienced, this might be the issue as I had an SSD failure and was lucky to have enough to recover and then implemented mirroring across two SSDs to save for any future outage. Since then I have had more errors than ever before. Only writing this now gave me this thought.
@Slade_Swan just so you know, Roon does not support any installations of Roon on Docker, so you might not get official support from them on this ticket.
Thanks. That may be the case but my question is largely agnostic of the particular install, at least as I see it. I have lost playlists and it would be nice to get them back if possible. If the upgrade of the database fails for any reason it would seem as if the previous backups from previous versions of the software are unusable (where that version required a database upgrade). Of course I might be completely wrong here however, there is no database upgrade process for a previous back up and reinstalling the roon_server per instructions (albeit from other install types) installs the current release. So if my hypothesis is correct you’re left with inoperable backups.
Do you by any chance have the Logs from the time of failure? You can try to check for those for clues to discover the corruption type, but it won’t really help with restoring a partial backup, unfortunately there is no way to do so.
Yes, SSD failure is a possible cause of database corruption. It’s possible the failing drive also is causing issues with the stability of the other backups.
Thanks for the reply. I will see if those logs provide any insight. Just to correct the interpretation of my previous statement.
In November/December 2023 I had a proper SSD failure highlighted by a synology alert. I recovered from this and implemented two new SSDs in RAID 1 configuration to save from future issues. However, since then I have had random database corruptions for which every time a database restore has recovered. This was until the most recent upgrade that required a database upgrade. This update worked but the database was not upgraded successfully and would not load similar to other occasions where the database had random failed. However on this occasion the previous database backups would also not load.
I assume this maybe because there is a one-shot database upgrade process and that there is no option of upgrading a previous database backup.
So my question were (even if only implicit):
Is my assumption correct regarding not being able to upgrade previous backups? If not could this be a feature request please as there is the opportunity to have a unrecoverable circumstance like I experienced.
Is there any chance that the RAID 1 on the disk containing the database is interfering and causing database corruption, bearing in that the SSDs are new NAS rated drives.
Older Roon backups should work with newer versions of Roon, it is very strange that you had this error.
Running Roon in a RAID 1 configuration is not something that we test. If possible, can you try to set Roon to run on a non-RAID partition to see if the issue reproduces there?
Were you able to locate the logs? Or reproduce the issue to get a new set of logs for the restore failure? If so, you can upload them to the below link and let us know, we can take a look at the specific error: