Active threads

Core Problem: Drives Stuck in PS0 at Idle drives remain in Power State (PS): 0 (4.5W consumption) and get burning hot in idle. While under load, copying 500gb of video files, temps settled at 35-37C. --- All three of your NVMe drives (nvme0, nvme1, nvme2) are detected by the kernel. They are identified as Viper VP4300L 4TB drives. They are correctly initialized and made operational within your md0 RAID array. The "clearing duplicate IDs" and "use of /dev/disk/by-id/ may cause data corruption" warnings are sometimes seen with certain NVMe controllers or firmware when used in RAID/LVM setups. They are typically benign if your system is otherwise stable and not showing data corruption, but it's good to be aware of them. These warnings are not related to your power state/temperature issue. Addressing modinfo: command not found The modinfo command is part of the kmod package on Debian-based systems like OpenMediaVault. It's not always installed by default in minimal environments. Please install it: Bash sudo apt update sudo apt install kmod Once kmod is installed, re-run: Bash modinfo nvme_core | grep default_ps_max_latency_us This will confirm the parameter's default value and any overridden values for the nvme_core module itself. Back to the Core Problem: Drives Stuck in PS0 at Idle Even with nvme_core.default_ps_max_latency_us explicitly showing 100000 from cat /sys/module/nvme_core/parameters/default_ps_max_latency_us, and the kernel reporting "No Workload," your drives remain in Power State (PS): 0 (4.5W consumption) and get burning hot. This is the central anomaly. Here's the refined plan forward: Re-confirm nvme set-feature Behavior (Crucial Diagnostic): This is the most important test right now. It will tell us if the drives are even capable of entering lower power states when explicitly commanded, or if something immediately wakes them up. Run this command for one of your drives when it's idle and hot: Bash sudo nvme set-feature /dev/nvme0n1 -f 0x02 -v 3 # Try setting to PS3 (value 3) Wait 30-60 seconds. Then immediately check its power state again: Bash sudo nvme get-feature /dev/nvme0n1 -f 0x02 -H If it changes to PS: 3 (or 4), even briefly: This means the drive can be told to sleep. The problem is then something preventing APST (Autonomous Power State Transition), likely constant, tiny I/O. If it stays PS: 0 or immediately reverts: This is more serious. It implies either a very aggressive and constant wake-up call, a firmware bug preventing power states, or a platform-level issue preventing the command from taking effect. Detailed I/O Monitoring with atop -d: This is your best tool to catch any background activity. Install atop: Bash sudo apt install atop Run it during an idle period when the drives are hot: Bash sudo atop -d Let it run for 5-10 minutes. Look at the DSK section for your NVMe devices (e.g., nvme0n1, nvme1n1, nvme2n1, md0). Pay close attention to RDDSK (reads/second) and WRDSK (writes/second), even very small numbers. If these are consistently non-zero, even if tiny, they might be enough to prevent deep sleep. Active Cooling is Highly Recommended (Urgent for drive longevity): Regardless of the software fix, 65°C idle is unhealthy for NVMe drives and will shorten their lifespan. Since there's no active cooling and ambient is 26°C, the drives are likely hitting their thermal limits during idle operation. Even a small 40mm or 60mm USB-powered fan aimed at the general area of the NVMe drives (or mounted in a case if possible) can make a significant difference. This is crucial for both diagnosing if cooling helps, and for protecting your drives. Consider NVMe Firmware Updates (If other options fail): If the set-feature command doesn't work, and atop shows truly zero I/O, then a firmware bug in the Viper VP4300L drives is a strong possibility. Check the manufacturer's website for firmware updates for that specific model. Let's get modinfo installed and then proceed with the nvme set-feature test and atop -d monitoring. That will give us the most actionable data.

Gnome uses Wayland and there some features don't work ... switch to X, where you will loose some functions, but remote desktop will work. FYI: https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

Hello. Regarding a tv box, I briefly tried to connect a Realtek 8188 USB wifi adapter and get it working, but I quickly quit trying because there wasn't any appropriate firmware in Armbian or there was some other technical problem, and I didn't know to solve the problem. I know for certain that wifi adapter works in Linux, for example on a PC and in EmuElec. I tried only XFCE4. But yes, I can recommend LXDE and LXQt as an alternative to XFCE4 because they are lighter. ☺

@tasknodes as SteeMan requested, you should provide details and steps of how you installed Armbian. also provide details of what caused a boot failure and details of what error messages or logs that you saw on your tv/monitor.

I have installed Armbian community images several times on different tv boxes that have an Amlogic S905x3, and usually there was no sound ♫. I wanted to get sound out of the tv box. So, I purchased a generic USB sound card audio dongle from Aliexpress for about $2 to $3 USD. (see attached photo). The USB audio device is described in Linux as a Texas Instruments PCM2902 Audio Codec (USB PnP Sound Device device 0 USB Audio), and when you run the command lsusb you will see the id numbers 08bb:2902. However, the specific hardware chip in the device that you receive may be different. I updated the databases for repositories by running sudo apt update -y and installed necessary packages to support the USB audio device and to use a graphical desktop such as XFCE4, for example: sudo apt install -y smplayer alsa-base alsa-oss alsa-tools alsa-utils alsamixergui pulseaudio pavucontrol pavumeter You can use smplayer or vlc or another as long as it is modern media player or audio player. Now make a backup copy of these two text files /etc/modules and /etc/modprobe.d/alsa-base.conf and save the backup copies in your home directory. Without quotes, edit the text file /etc/modules and write “snd-usb-audio” at the end of the file. Also, edit the text file /etc/modprobe.d/alsa-base.conf and change a few things. Put a hash # symbol in front of every “options snd-usb-audio index=-2” that you see in the file /etc/modprobe.d/alsa-base.conf . I found two of them. Add these two lines of text to the bottom of /etc/modprobe.d/alsa-base.conf without quotes. "alias snd-card-0 snd-usb-audio" "options snd-usb-audio index=0" Make sure that your user is in the 'audio' group. Run the usermod command against a regular user (not root) that you want to use often. And reboot. sudo usermod -a G audio {username here} sudo reboot Login to the graphical desktop, open the volume control program by clicking on the panel Applications, Multimedia, PulseAudio Volume Control. Click the Configuration tab at the far right of the volume control program and change the profile to Analog Stereo. * Important – now connect a headset, earphone, or amplifier to the 3.5mm (⅛’’) 🎧 headphone receptacle on the USB audio dongle and play some sounds, music or youtube. You can use wav audio files in the directory /usr/share/sounds/alsa/ if you don't have an internet connection.

thanks @Igor , this is really fresh like an hour ago would check in the rolling releases for images , it is deemed more stable for dram detection , sizing at boot. there tend to be dram size detection and other issues which I'm not too sure if that might be a timing related issue.

Personally, I would suggest to not even run a Desktop Environment, certainly not on a true IoT device. What GUI software do you plan to run? Are you aware that X can forward a GUI over ssh for you? Let's say you wanted to run blender and display the output at home or in your office on your shiny 27-inch, curved 4K display. OK. First, install the software on the IoT device "sudo apt install blender". Secondly, make sure X Forwarding is enabled on the IoT device "sudo grep X11Forwarding /etc/ssh/sshd_config" and make sure it is set to yes. Then, you should be having no problem to get blender to run on your IoT device but the GUI window will show on your large display with "ssh -X $IP-of-your-IoT-device", run this of course from your beefy machine. This works just the same for any other software you want to run remotely.

Have you tried to ask this in our discord server?

I'm sorry, but I don't see where the problem is. Please be more specific.

it would be good to have those logs for a non-working setup.

awesome, well done! and thank you for sharing.

check those files /usr/lib/systemd/system/armbian-resize-filesystem.service /usr/lib/armbian/armbian-resize-filesystem

Please someone help I can't start X11... I installed: sudo apt install xserver-xorg-video-fbdev openbox x11-xserver-utils xinit xserver-xorg-input-evdev Created this /etc/X11/xorg.conf.d/99-fbdev.conf file And tried sudo startx, then the LCD got a black screen for a fraction of a second, and I got (in my ssh session) The error message is FBIOPUT_VSCREENINFO: Invalid argument I also see that xorg wants to use 24 bit color, but my LCD is 18 bit color. I will try to get X11 working again with the linux 6.13 build folders that worked for me in april

Here are a few generic tips to consider: - Going to something industrial grade is always good. mean-well would be entry-level here - Phone chargers are not good for powering SBCs. They're designed to handle constant load while SBCs draw is variable depending on load. - Usually SBCs prefer slight overvoltage. So it is usually not an issue to feed them with an PSU that outputs 5.1 or 5.2 volts since this compensates for losses across wires, connectors and the PCB anyway. I personally have a mean well HRP-200-5 which powers all my sbcs but would be way too much for a single one. Official chargers from Raspberry or Orangepi would do too since they already come with slight overvoltage by design.

# dpkg -L linux-u-boot-nanopi-r6c-current | grep platform_install.sh /usr/lib/u-boot/platform_install.sh that script writes the uboot binary

Armbian doesn't support TV Boxes. Read the following to understand the status of Armbian and TV Boxes:

Most SBC backup tools I have seen are based on rsync and they operate then on/with Ext4 filesystem. Is fine when you don't have complex servers and databases running. I am not sure about pihole, but I saw that by default it keeps a 1 year history so a database file of about 1GB. I am not sure what happens to integrity of such a file on the target if during copy/rsync the source file also changes. AFAIK databases have their ways to handle it, when power-loss or so, but not sure. If you want to be sure, use Btrfs as filesystem, then you can make atomic snapshots and use those for source of rsync or use: btrfs send -p <old snapshotnumber> <snapshot number>| ssh <remote_host> btrfs receive <backup folder>. That is the basic manual option. Complete tool is btrbk, it is standard in Debian repo, see https://digint.ch/btrbk/index.html docs of author. You still need some partition and bootloader handling I think. Or assume that a total crash is rare so only backup to NAS or so and reconstruct manually if you need a new SD-card or so or want to copy thing so eMMC of the OPi3.

Thanks for the info and the PR link. So the fix is in the works but not merged yet — any idea how long it might take to land in nightlies or mainline? Should I keep the thread open until then, or just close it and reopen if needed after testing? Thanks again.

Ver 20250627-edge kernel 6.16-rc3

Armbian doesn't release this old LTS release now, but you can build one from github.com/armbian/build Yes, flatpak version is using ffmpeg in its own sandbox without rkmpp support, if you build the deb locally, it will get linked with ffmpeg on system which has rkmpp support.

I am a new user to Allwinner T527 SOC in OrangePi 4A.... The SBC is great for general purposes but VPU really sucks. On the other hand my RK3566 SOC in OPi3b & Radxa Zero 3E shines when using JellyFin or Chrome

They are in the same pin numbers, but different GPIO Update: I found that I can just use both CS0 and CS1 in opiz3 normally as expected in a DTS for a SPI ili948x LCD and touch screen, as seen in my ili9488 thread. (PH9=CS0, PC10=CS1) I was having so much trouble before, but now I finally got it

Since this issue is well over a year old, I'm not sure where you stand with this. Are you still looking to get a fix? If so, answering all of the questions asked would be helpful. When you say "broke again" what do you mean by broke? Apparently the box is running and you can update it as you continue to update it per the message above saying you have upgraded to 6.12.32. Have you tried a fresh image installed on a new SD card to see if that works (vs. your upgraded system)?

Thank you - I should have checked w/ df. The reason is that /var/log is mapped to /dev/zram1 which is 47M for my system and auditd.conf requested space_left=75 (MiB, that is). I changed this to space_left=25% and admin_space_left=5%. No it starts to log something at least.

Technically a DTS is specific to a kernel version. As changes happen to the kernel apis, those need to be reflected in the DTS's. So they DTB and kernel image are always built and shipped together. Now in practice changes that impact the DTS aren't common, so you can usually use DTS's accross kernel versions. However in this case, I'm assuming that when you refer to 6.1 that is a vendor kernel, and 6.10 or greater would be mainline kernel. Now these are two very different code bases. A vendor kernel is whatever the board/cpu vendor hacked together to get a kernel running on their hardware. Often these are more android kernel based. When features finally get rolled into mainline, their implementation is often completely different. Thus one can't share a DTS between a vendor kernel and a mainline kernel. But often you can use the vendor DTS to gather information about what needs to be implemented in the mainline DTS.