· None of the above quite fits
· None of these quite match
· RoonServer stuck "Initializing" / web UI hangs on logo — fresh database also hangs
· Araknis AN-310-RT-4L2W router + AN-110 switches, UniFi APs managed separately via CloudKey
My ROCK install (Intel NUC) will not finish starting RoonServer. The web UI loads the logo and hangs indefinitely, and Roon Remote shows the server as “Ready” but tapping Connect does nothing, or cycles back to “Waiting for your Roon Server.” A fresh RoonServer folder hangs the same way as the old one, which is what’s confusing me.
Intel NUC running ROCK (RoonOS 1.0 build 259, Roon Server 2.66 build 1658)
OS/system drive: M.2 NGFF SATA SSD (ZTC 128GB)
Music library stored locally on the NUC internal storage (ROONDATA / Storage)
Network: Araknis AN-310-RT-4L2W router + AN-110 switches, UniFi APs managed separately via CloudKey
Server has a DHCP reservation at 192.168.2.144
The whole thing began as a network issue — connecting the Roon server via Ethernet appeared to cause a broadcast storm (switch LEDs going crazy, whole network dropping). After a lot of diagnosis that turned out to be a separate rabbit hole, and in the course of it the NUC was moved to a different switch port, lost its IP reservation, and went through multiple power cycles. Since then RoonServer has not come back up.
Confirmed the NUC is powered and the drive activity light flashes (roughly once every 10 seconds now — looks idle, not actively rebuilding).
Re-established the DHCP reservation at 192.168.2.144 and confirmed the server gets that IP.
Power cycled the NUC many times, including a hard shutdown via the power button. Waited well over 10–30 minutes on multiple occasions.
Confirmed the web UI at http://192.168.2.144 loads the Roon logo but never progresses.
Force-closed and reinstalled the Roon Remote app. Server appears as “Ready” but Connect is unresponsive.
Pulled the SATA SSD, mounted it on a Mac (Paragon extFS) and browsed ROONDATA successfully.
Checked RoonServer/Logs — the Logs folder was empty.
Copied the entire RoonServer folder to my Mac as a backup, then renamed the original to RoonServer_old so ROCK would build a clean one.
Put the drive back in — ROCK came up “fresh like new” and prompted for an update. I accepted the update.
After the update, the server again will not connect and the web UI hangs on the logo exactly as before — even though this is now a brand-new RoonServer folder.
A fresh RoonServer database hanging the same way as the old one suggests the problem isn’t database corruption. What else causes the web UI to hang on the logo at boot?
Could the update have been interrupted and left ROCK in a half-updated state? Is there a clean way to recover/re-flash without wiping the local music library on ROONDATA?
Could this be the SATA SSD failing? Any way to check drive health given ROCK has no shell access? (I have the drive in a USB enclosure and can run SMART from a Mac/PC.)
I have my old RoonServer folder safely copied off. If I get a clean boot, what’s the correct way to restore the old database/settings from that copy?
Playlists, play history, library edits, audio zone settings (the contents of the old RoonServer folder)
My local music files (these live in Storage and I believe they’re untouched)
Any guidance on the most surgical recovery path would be hugely appreciated. Happy to share specific logs or diagnostics if someone tells me which file to pull off the mounted drive.
Follow-up with additional diagnostics:
SSD health: Tested in CrystalDiskInfo — Health Status “Good,” 0 reallocated sectors, 0 read errors. The drive hardware appears fine. Notable: 145 unsafe shutdown count across 201 power cycles (a lot of abrupt power loss during troubleshooting).
Codecs folder: Populated (67.84 MB) — ffmpeg/codecs appear correctly installed.
systemlogs folder: Empty (0 KB) — this seems significant. The server doesn’t appear to be writing any boot logs.
RoonServer/Logs: Also empty.
Old database: Preserved as RoonServer_old (28.4 GB) and also separately copied off the drive onto another machine, so my settings/playlists are safe. Happy to restore from it once the server boots cleanly.
Fresh RoonServer folder (1.36 GB) was auto-created on boot but hangs identically to the old one.
A filesystem recovery scan flagged ~2.66 GB of orphaned/deleted data, which — along with the unsafe shutdown count — makes me wonder whether the ext4 filesystem took some damage.
Access for diagnostics: I can pull the SATA SSD and mount it read/write on a Mac (Paragon extFS), so if there’s a specific log or file you’d like me to retrieve or inspect, just point me to the path.
Given the empty systemlogs and the fact that a fresh database hangs the same way as the old one — could the OS update have been interrupted mid-write and left ROCK in a half-flashed state? Is there a safe reflash/reinstall path that preserves the local library in Storage?
RoonOS is healthy again, but running on a fresh database — how do I safely restore my original RoonServer database (I have it as RoonServer_old and a full file copy) without a Roon backup file?
Update — resolved. Two things were going on. First, the server wasn’t getting a stable network identity; setting a DHCP reservation and confirming a clean wired connection got RoonOS booting and advertising normally again. Second, the database was in a bad state after an interrupted Server upgrade. With Roon Server stopped via the web UI, I swapped the database folder over SMB (renamed the fresh RoonServer aside, renamed my original folder back to RoonServer, confirmed Cache/Core/Database/Registry/Settings were directly inside). On restart everything came back — all 1,117 albums, playlists, favorites, history intact.