Intermittent playback dropouts with Linn Majik DS via RAAT (ref#81C2UI)

Support
1

Hi! What’s not quite right with Roon?

· None of the above quite fits

None of the above quite fits

· None of these quite match

Tell us what's going on

· Subject:
Intermittent playback dropouts with Linn Majik DS (RAAT only)
Description:
Hello,
I have intermittent playback interruptions when using a Linn Majik DS with Roon.
System
Roon Server on Synology DS218+
Linn Majik DS connected via Ethernet
TP-Link TL-SG1005D switch
Local music library on Synology and TIDAL
WiiM streamer on the same network
Problem
Playback occasionally stops briefly and then skips forward. It is not a click or pop.
The issue occurs with:
local files
TIDAL
44.1 kHz
96 kHz
192 kHz
The interruptions are random and can happen immediately after playback starts.
What has already been checked
Linn firmware up to date
Roon up to date
Ethernet cable replaced
Linn power-cycled
Synology CPU load normal
DSP downsampling reduced the frequency of dropouts but did not eliminate them
Important observation
The same Linn plays 192 kHz WAV files flawlessly using Linn Kazoo (OpenHome/DLNA).
The WiiM also plays through Roon with almost no interruptions.
This makes me suspect a RAAT-specific issue rather than a general network or hardware problem.
Could you please advise how to diagnose this further or which logs you would like me to provide?
Thank you.

Tell us about your home network

· AVM FRITZ!Box (latest FRITZ!OS), TP-Link TL-SG1005D Gigabit unmanaged switch. Roon Server is running on a Synology DS218+. Linn Majik DS and WiiM are connected via Ethernet. iPads/iPhones are connected via Wi-Fi. No VPN is in use. Ethernet cable to the Linn has already been replaced.

2

Hey @holger.wiedemann,

Thanks for writing in and for sharing your report. Welcome to the community as well! We’re happy to have you here, but also apologize in that it’s an issue that has prompted your first post.

We were able to review a fresh set of Roon Server logs, and they told a clear story, and it points somewhere a bit different from where we’d expect.

First, a quick clarification on the setup, the Linn Majik DS isn’t actually playing via RAAT. Linn devices are OpenHome endpoints, so Roon streams to them using Songcast. In your system, the RAAT device is the WiiM. That matters because the logs show the interruptions aren’t specific to either protocol, they’re happening on both endpoints.

The Linn logs 256 “time discontinuity” events (the audio timeline jumping forward, which matches the brief stop-then-skip you’re hearing), and the WiiM logs nearly 500 dropout events too. It recovers more gracefully thanks to a larger buffer, so it feels smoother, but it’s affected by the same underlying cause.

That cause is on the server side. The Roon Server process on your DS218+ is experiencing very large garbage-collection pauses, moments where the whole process freezes. Across the logs, more than a fifth of the time windows show pauses over 2 seconds, and a number of them exceed 10–15 seconds. The dropouts line up almost exactly with these freezes: whenever the server stalls, it can’t deliver the real-time stream on time, so the endpoint gaps and skips forward to catch up.

The reason it’s this severe is memory and CPU load. Your library is around 60,000 tracks, which pushes Roon’s memory use up near 3 GB, and the DS218+'s processor simply can’t keep up with managing a heap that large while also streaming. That’s also why Kazoo plays the same 192 kHz files flawlessly: with Kazoo the Linn pulls and buffers the file itself, so a brief server stall doesn’t matter. Roon (both Songcast and RAAT) sends a continuously clocked stream, which a multi-second freeze breaks immediately.

So the good news is this isn’t your Linn, your cable, your switch, or your network, those are all fine. It’s the server being under-resourced for the library size. A few things to try, roughly in order of impact:

  1. The real fix is moving Roon Server to more capable hardware. An Intel NUC-class machine or a Nucleus will eliminate this entirely. The DS218+ is below what this library size comfortably needs.
  2. If you'd like to stay on the Synology for now, please confirm the NAS RAM is upgraded to its 6 GB maximum, and check that nothing else heavy (Docker, Plex, indexing, etc.) is running alongside Roon.
  3. In Roon settings, reduce background load by setting audio analysis to a lower or throttled speed so it isn't competing with playback.
If you make a change, keep the same logs running and we can confirm from them that the GC pauses have dropped and the dropouts are gone.

Happy to help interpret the results once you’ve tried one of the above. :+1:

3

Thank you very much for taking the time to analyse the logs in such detail. I really appreciate the thorough explanation.

Your analysis of the GC pauses makes sense, and I understand how these could lead to the dropouts I’m experiencing.

However, I have one question regarding the conclusion that the DS218+ is now underpowered. The issue did not exist in the past with essentially the same music library size and the same Synology DS218+. My library has not grown significantly since then, and the hardware configuration has remained unchanged.

If the hardware is now considered insufficient, what has changed? Could a recent Roon update, a database issue, or another software change explain the newly observed long GC pauses, rather than the hardware itself?

I’d be very interested in your thoughts on this, as understanding why the behaviour has changed is important before I consider moving my Roon Server to different hardware.

Thank you again for your support.

4

Hey @holger.wiedemann,

That’s a fair question, and an important one to answer before you go buying hardware, so I’m glad you asked rather than just taking my word for it.

The short version: you’re right to suspect software rather than your hardware suddenly getting worse. The DS218+ hasn’t changed, and neither has your library in any meaningful way. What’s changed is Roon itself. Over the past year or so, as the app has evolved. On a machine with plenty of headroom that goes unnoticed, but on a RAM-constrained NAS running a ~60,000-track library, it’s enough to push the process past the point where the garbage collector starts producing the long pauses we saw in your logs. So it’s less that your hardware became insufficient and more that the software gradually grew into the ceiling your hardware always had.

Here’s the genuinely good news, and the reason I’d hold off on hardware for a moment: this is exactly what our team has been working on. There’s an Early Access build out right now (Roon 2.71, Build 1674) specifically targeting server memory usage, and the early results are promising, managed memory is holding much flatter than before. That’s precisely the behavior behind your GC pauses, so it’s very much worth testing on your setup before spending anything.

If you’re open to it, I’d suggest:

  1. Head to our [url=https://help.roonlabs.com/portal/en/kb/articles/roon-early-access-program]Early Access KB article[/url] and download the Early Access build for your platform, then install it over your current Roon Server and Roon remote devices.
  2. Give it a few normal listening sessions. See how things perform.

If you’d rather wait, these same updates will eventually be merged into our next Production update, which will auto-prompt you in app, like usual.

One honest caveat: even with these improvements, a library your size on a DS218+ is close to the edge, so the EA build may resolve it fully or may just reduce it, and that’s genuinely useful for us to know either way. Moving to more capable hardware stays the guaranteed long-term fix, but let’s see what the new build does for you first. :folded_hands:

5

Hey @holger.wiedemann,

Wanted to follow up on this, and we’d still like to see how the Early Access build is behaving on your DS218+ before you make any hardware changes. Have you had a chance to install Early Access on Roon Server and your Roon Remote devices, then use it through a few normal listening sessions to see whether the long GC pauses have eased up? If you already tried it, let us know what changed, even if the improvement was only partial.

Please note, if we don’t hear back from you this thread may close automatically soon. If the thread auto-closes and you need further assistance, please submit a reopen support request via the technical support help form below and specify that the issue should be reopened. Thank you.