I’m happy to share log files, just tell me where to find them and how send them.
@HeavyMetalDrummer take a look at https://help.roonlabs.com/portal/en/kb/articles/logs for directions and visuals.
I don’t understand how I have zero of those issues by simply running roon on a headless M4 (base) Mac Mini that is currently even only connected trough WiFi . Sometimes no reboots for months and no reboots after a roon update.
External library connected through thunderbolt4 drive with a nvme ssd.
Maybe time to abandon the underpowered outdated Nucs ? Or at least enable swap to a internal nvme (not that I ever have my m4 seen using swap)
Hi all, no issues with my setup before and after the recent upgrade, 2191 Artists, 4425 Artists, 4,9042 Tracks. Silent Angel Z1C, 8gb ram. Music stored on 4 Tb SSD.
I have an average NUC and none of these problems either. It’s not that. (It’s also nice not to have to worry about abandoned OS versions like the people who can’t update to macOS 12 now
)
God, why did I spend an hour to explain memory management in Roon OS.
Exactly the same process happened here with my Rock Nuc. Great progress with 2.65 and after one day or two waiting and waiting again, reboot and speed again. I add: I live in Argentina, and early in the morning, when America still sleeps everything is ultra fast adn precise, during the day and mainly in the evening Roon becomes slower and slower. Cheers. Diego.
My server logs are flooded with these:
May 01 13:39:35 prince start.sh[1476585]: System.Net.Sockets.SocketException (104): Connection reset by peer
May 01 13:39:35 prince start.sh[1476585]: at System.Net.Sockets.Socket.AwaitableSocketAsyncEventArgs.ThrowException(SocketError error, CancellationToken cancellationToken)
May 01 13:39:35 prince start.sh[1476585]: at System.Threading.Tasks.ValueTask`1.ValueTaskSourceAsTask.<>c.<.cctor>b__4_0(Object state)
@Stephen I just uploaded my log files.
The sound is worse than before, what happened?
That’s not possible with bit perfect playback
Most likely, nothing changed in Roon. There are “better!” “worse!” claims after every big update.
If something changed for you, check your system or ears.
Swap is so… twentieth century.
TBH, it’s only used to avoid OOM errors and crashes… I think I read about this happening somewhere. ![]()
My NUC 10 has been running smoothly for 4 years. No problem at all. It was fine before the latest upgrade and it is up and running after it.
Hi @spockfish - can you upload your logs so we can take a look please?
Thanks @HeavyMetalDrummer - I’m on it.
Same here. I’m running a Nucleus with 8 GB RAM, about 1,000 local albums and ~3,200 tracks. Like with previous releases, the main page gradually gets slower to load, I end up staring at the jellyfish for 3-7 seconds on average.
After 4-6 days, it gets to the point where clicking the play button does nothing, and a reboot becomes necessary.
I even made a maiden installation from scratch on the Nucelus and reinstalled a new dB and Roon remote.
B/R Stefan
Not appropriate for Rock users, but on my M4 Mac Mini Pro I just have Roon set to restart every morning using crontab, the same way I did on DietPi before that and I make sure I do it about an hour after the processing is supposed to stop.
Seems to work great for the 4 months I have been doing this
Maybe it should not be necessary, but it always has been in my experience
Rarely look at this forum but I have noticed ROON is slow on my MacBook yet is super fast when I use an iPad. My core is located on an Apple M4 mini Mac, library size 3,300 album. I will now watch this thread with interest.
I let Claude analyze my logs. Here‘s some insights:
1. Not a classic resource exhaustion
-
Memory stable (~1.3 GB physical)
-
GC negligible (~0.78%)
-
Thread/handle counts stable
→ This rules out the usual “it just runs out of RAM/CPU”.
What I now will do, is restart the server and after one day save the logs. Wait until I have to reboot again and then let the latest log to be compared with the firtst one
If any of you having performance problems did a database backup & restore as part of your recent update process (for instance, moving from Roon in NAS to the Docker package) it does appear that Roon needs several several hours of background processing to revisit your collection even with what appears to be a perfect database restore. Setting up a couple of 12+ hour overnight scheduled background processing sessions in Settings > Library without rebooting has improved things for me. Not yet perfect, but much better. I’m still experimenting a bit. I’ll keep folks posted with what I learn.