DietPi - Roon server - Some questions

It’s definitely a strange one, and my device on the same kernel version is fine. But I am not backing up to a Windows device. One Local backup on DietPi to a 3rd Hard disk and 2 different Synology backup’s, but they all work fine for me

1 Like

Which SMB protocol versions are at play here? Windows does disable v1.0 by default (could be enabled manually though not recommended).

Well my Synology only had SMB 2 and 3 enabled to it and that works, so that does suggest that Roon is happy to talk over that protocol.

Personally I am thinking this is some funny after Windows 11 24H2 update was installed, but this is purely a hunch

I don’t know what is going on but, as an experiment I have just created a share on my laptop recently upgraded from Win 11 23H2 to 24H2 with no policy editor changes or registry changes.

Rather than use ‘Everyone’ to allow all users read/write permissions, I left ‘everyone’ with read only permissions and then added my windows user name as an additional permitted user with full control.

Then, on my DietPi Roon Server, I used dietpi-drive_manager to mount that share under /mnt/wade

Looking at the owner and permissions on that folder, the owner was dietpi:dietpi and the permissions were ‘drwxrwx—’.

Using the an ssh terminal on dietpi, I was then able to successfully do:

cd /mnt/wade
echo hello wade > wade.txt
cat wade.txt

The last command correctly displayed “hello wade” as you would expect and I was able to see the newly created file on my laptop and verify the contents again.

Like @Torben_Rick and @Michael_Harris, I am running the 6.12 kernel from debian backports:

uname -a
Linux Roon 6.12.9+bpo-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.12.9-1~bpo12+1 (2025-01-19) x86_64 GNU/Linux

Note: I use a local user on my Windows 11 laptop - not a Microsoft account - with that user name and the login password being used as the credentials in dietpi-drive_manager when mounting the share.

2 Likes

I’ve been using DietPi for Roon Bridge endpoints for ages, and skimming this tread has me tempted to try DietPi for Roon Server. Would DietPi give improved performance (loading of pages, searching, loading of play queue, etc.) compared to Windows 11 on the same hardware?

My current server is a Dell Optiplex 5090 SFF with 10th gen i5, 16gb, NVMe 512gb OS drive. I use DSP (parametric EQ or convolution) on most endpoints, but rarely have more than one going at a time.

Also, my system is idle probably 75% of the time. Would DietPi consume less energy at idle?

Thanks for the insight!

Andy I don’t think it would make a noticeable difference to performance in truth. Most of us run it for the flexibility it gives us to run other services like Plex, LMS etc

As to power savings it is possible to run it in low power mode, but I found that a bit slow as it kept the processor at a low maximum setting

4 Likes

I tend to concur with @Michael_Harris on this one. I moved from Windows to DietPi because I had a fanless PC that I couldn’t upgrade to Windows 11. Roon on DietPi is rock solid, but so was Roon on Windows.

4 Likes

I also feel that moving from a Windows hosted Roon Server setup to Dietpi hosted on the same hardware is unlikely to yield any noticeable performance improvement.

The only time this might not hold true is if system is memory limited in which case, moving to a lighter weight DietPi may give a performance improvement because of reduced paging. However, since your Windows computer has 16GByte of RAM installed, I would imagine that you would have to have a very large Roon library before this became an issue.

With regard to energy consumption:

Like @Michael_Harris, I found the DietPi low power mode to be a bit too low power on my NUC11TNHi7. However, I did find that ‘performance’ mode with Turbo disabled worked quite well for me whilst I was only running Roon Server. It should be noted that my Roon library is quite small.

The Dell Optiplex 5090 SFF was available with three different 10th gen i5 processors - the 10400, 10500 and 10505. These are desktop processors and, as a result, are likely higher performance than my NUC. Consequently, it is more likely that you will get acceptable performance without turbo enabled on your i5 - especially if it is the 10500 or 10505.

Disabling Turbo mode reduces the maximum performance of your computer (and thus saves energy) only when doing compute heavy tasks (which I don’t think you are doing very often) and so I would imagine that any power saving would be absolutely minimal.

Having said that, this is not necessarily an advantage of DietPi. You may be able to achieve the same on Windows by disabling the turbo mode in the system BIOS.

2 Likes

@Wade_Oram - I have a “small” challenge related to energy consumption.

If I turn off my DietPi headless PC than the new OWC Express 1M2 (see below) is still one. But I can’t finde any setting for that.

That is not the case when connecting OWC Express 1M2 to a Windows 11 PC. If I turn-off the Windows 11 PC than OWC Express 1M2 also turns-off

Torben

You are sure the DietPi is actually off when you think it is?

I would imagine that is either a machine architecture or BIOS setting thing. I know you are using Thunderbolt (rather than plain USB3/4) but I have had PC’s in the past which supply permanent power to the USB ports (or at least a subset of them).

It is certainly nothing to do with the DietPi performance settings which are related to the processor only.

In the photo the round power LED is lit around the power button

It could be a USB port that still has power when the OS is shutdown

SSH into DietPi if no screen is attached and try

shutdown now

Does it still stay powered?

On the picture both are on :slight_smile: If was just for illustration

Here a real PIC where the DietPi PC is off and the Thunderbolt is still on:

Torben