Having a repeated problem since 943 updated that during a complete scan our QNap TBS-464 on 5.0.0.1986 reboots during the scan. No errors in the Qnap logs, I just see the device restarting.
Due to the other big bug with Roon on QNap with scanning, running a complete scan is required whenever Roon or the Qnap restarts, so we have a dead Roon now as whilst the Roon app is started it won’t complete a scan.
Really disappointed with this. Having been stable for years I feel Roon has now entered the world of software quality more akin to Naim.
If you are willing to look into the Roon logs as well, it would be interesting to see if there are any files right before the reboot that Roon gets stuck on analyzing.
Note there is something weird in the zip file the dump created. It is 108MBs in size and similar size when extracted. But when that extract is recompressed is 8MB.
From my view of the logs, I don’t see it starting to scan.
With 943 it was almost instantaneous the moment the scan started. 952 gets about 1/3 of the way through a large-ish library, around 100,000 tracks, before restarting.
The new database location makes no change to the behaviour.
This is an NVMe SSD NAS so the scanning is super fast if that could be a factor, but it was running and stable with the previous release to 943. The previous setup was regular SSD for the db and WD Red hard drives and that was stable for years.
Without Roon running a scan the device is stable. The only other running apps are Asset and a QNap HBS sync job to DropBox which is not actively syncing content at the time of the scan.
Can you please try to move this directory out of your Roon watched location and verify if the same issue persists? It looks like Roon was scanning this directory right before the crash occurred: /DropMusic/Music_Mum&JacksCDs/London Filmworks Orchestra - Colin Towns/The Beatrix Potter Music Collection/
I am unsure why your QNAP would fully restart due to a Roon issue, if you have not yet performed a hard disk integrity check, I would do so, as this if the first case I’m hearing of a reboot during a scan.
I did take another look over your logs and I believe I found a crash, the last thing that Roon was doing was performing analysis on this file: /DropMusic/Music_Downloaded/_Qobuz/Maria Callas - Remastered The Complete Studio Recordings/ and importing this album: /DropMusic/Music_Downloaded/_Qobuz/Prince-Batman/.
If removing these two folders does not resolve the crash, then I would suggest temporarily disabling the current storage and only leaving a handful of folders enabled, to confirm that the system is stable without the full library.
The reason the QNAp would restart whilst running Roon is due to an issue with the scanning process.
Thie system had been completely stable, QNap file/folder permission bug aside, since installation. It ran one or multiple of the 930 releases without issue. This issue only occurred with and since build 943.
The Qnap has no issues with Asset running and is 100pc stable, except when running the Roon scan.
Running a Roon scan on the folder with say ten albums completes and the unit is stable post-scan. It is only when the scan gets to somewhere around the high hundreds of albums that it then restarts.
I have tried fresh installations with reboots of the Qnap between, with and without restoring a previous backup. The behaviour remains.
So whilst I appreciate you don’t understand what the problem is and look elsewhere for a solution, the above testing suggests to me that the problem lies in the scanning code of recent releases.
If you disagree, perhaps you can find a way for us to run a 930 build?
Thanks for that additional information. Since it is not clear which file(s) are causing the issue according to the logs, the best way forward would be through a sort of binary search, where you add half of your music files and check to see if it reproduces the issue.
If it does, then you know the offending album is part of that half, and if it doesn’t then you know the problematic album is part of the other half. Unfortunately, locating which album is causing the issue is not possible from your logs it seems. Would such a test be possible?
