I got tailscale and Plex all setup, bought a new router just to to run tailscale at my remote location, everything was sweet... for a week... then it inexplicably stopped working. Everything seemed fine... the NAS was accessible remotely, I could ping it, access it's admin interface. I could even access Plex on my computer and play movies. But on my TV and phones, nope. It was like my NAS was completely invisible... even to Remote access viewing, let along "local" viewing. I was tearing my hair out, eventually by chance looked though my NAS settings and found Remote Access "On" with a 100.x ip address that looked like a tailscale range, but wasn't its correct tailscale IP. So apparently this is Terramaster's remote access solution, and it takes an IP address range that tailscale thinks it owns, and this can lead to inconsistent behavior. So I disabled that, you don't need it I guess if you are using Tailscale. And hallelujah, everything started working again. So I post this for others to find... if you're using tailscale on your Terramaster, turn off Remote Access.

Hey everyone,

I wanted to share a major breakthrough I just had after months of hair-pulling, troubleshooting, and testing. If you are running TrueNAS SCALE on lower-power Intel Alder Lake-N95 (or N100) processors—specifically pre-built setups like the Terramaster F4-424—and experiencing random hourly/daily hard freezes or overnight lock-ups, this is highly likely your issue and your fix.

The Setup:

• NAS: Terramaster F4-424
• CPU: Intel N95 (4 Cores)
• OS: TrueNAS SCALE (v25.04.2.6)
• RAM: 16GB (Single stick—waiting out the current RAM pricing insanity before buying a second!)
• Storage: 2x NVMe drives (Boot + App-Data), 4x HDDs (Data Pool)

The Symptom:

The server would completely freeze up. Not just a web UI drop—total networking dead, SSH completely unresponsive, and a frozen screen if a monitor was plugged into it directly. There were absolutely zero error logs, core dumps, or kernel panics written to journalctl because the host processor was stalling silently, usually during low-utilization windows or overnight when the system sat completely idle.

The Problem:

The Intel N95 has a known hardware/firmware interaction bug on modern Linux kernels when transitioning into deep power-saving idle sleep states (specifically C6/C8/C10). When the CPU tries to drop its power consumption down or wake back up from these deep states, the internal voltage drops past a stable threshold, instantly locking the processor state machine.

A common workaround online suggests forcing the CPU scaling governor to Performance, but this forces the CPU clock cycles to stay pinned higher than necessary, wasting power and causing unnecessary heat when the NAS is sitting idle.

The Real Fix (C-State Limiting):

The true, precise engineering fix is to leave your system scaling governor on its optimized powersave profile, but explicitly tell the Linux kernel to limit how deep the CPU can sleep. Capping the idle states to C1E preserves perfect system stability while still allowing the CPU to scale its active clock frequencies all the way down during zero-load periods.

Because TrueNAS SCALE is an appliance, manual edits to /etc/default/grub can be wiped during standard system updates. The correct and safe way to apply this parameter permanently is using TrueNAS's built-in middlewared configuration utility via the shell.

How to apply it:

1. SSH into your TrueNAS box (or open the Web Shell) as admin and elevate to root (sudo -i).
2. Run this command to natively append the parameter to the TrueNAS kernel options database: Bashmidclt call system.advanced.update '{"kernel_extra_options": "intel_idle.max_cstate=1"}'
3. Regenerate your active GRUB boot configuration with the newly embedded setting: Bash/usr/local/bin/truenas-grub.py
4. Reboot your NAS.

How to verify it's working:

Once the system comes back up, you can run this loop in the shell to check your core idle stats:

Bash

for dir in /sys/devices/system/cpu/cpu0/cpuidle/state*/; do echo -n "$(basename $dir): $(cat ${dir}name) -> Usage: "; cat ${dir}usage; done

You should notice that your usage counts are concentrated entirely in state0 (POLL) and state1 (C1E), and the deeper, unstable states (state2/C6, etc.) will completely vanish from the loop.

Meanwhile, if you check your scaling governor, it will still report a safe and efficient powersave.

Since switching to this specific C-state limit, my server uptime has been absolutely pristine, and the overnight crashes have completely vanished without sacrificing the power efficiency of the Terramaster hardware. Hopefully, this saves someone else months of debugging! Let me know if you’ve run into this on your low-power boxes.

The other day I found out that my tailscale connection to plex wasn't working. When I tried to fix it I found that I couldn't and I eventually discovered that docker manager couldn't pull images anymore. At some point when trying to fix it I rebooted my router only to find after it rebooted my plex container was gone and I had to recreate it. I did that and now when trying to claim the server to connect my tnas (TOS 5) to plex it refuses to claim the server. I've spent the past 4 hours trying to figure this out. I've changed the DNS to [8.8.8.8](http://8.8.8.8) to resolve it. I've reinstalled plex several times, I've manually added the claim token in the container, I've tried using an incognito browser. It simply won't let me claim the sever. I don't even care about tailscale anymore I just want plex to work on my local network. Could anyone please offer some suggestions? This is really killing me.

So I recently picked up the F4-212 and updated it to TOS6 thinking it'd work well for me and from what I read I could use docker for things, and from what I've been seeing other things say to just go through Docker

However Docker isn't in my application centre? And the community site fails every download I try to do (And has a fun limit).

I'm trying to figure out if I'm missing something here or what because nothing I'm finding seems to give me an answer

Thanks to any direction, I'm utterly confused because I thought this stuff would just... Work?

I have the Terramaster D2-320 which is Direct Access Storage. I've had it for about a year and have had no issues. I've been using it for my media server. Configured it for RAID 1 with NTFS file system on my main Windows 11 computer.

I recently got an HP Prodesk and wanted to move my media server to that. It's also Windows 11. However, when I plug in the DAS it isn't detected as a drive. It shows up in Control Panel and Device Manager as a USB but it doesn't show up in file explorer as a drive. Nor is it in disk management.

If I plug it back into my primary computer it shows up without issue. Does anyone know what might be going on here?

Any help would be appreciated because I'm lost at this point. Thanks!

unraid has internal boot coming with a release candidate out. It allows unraid to boot from a SSD or NVME (inc. mirror). Has anyone tried this on a F4-424 (sorry wrong title) yet? I'm running the N95 model.