NOBL

My lsusb just shows the "ids" of the controller of connected drives but not the drives themselves. Neither the external disks nor the Ethernet HW were detected. I checked the directories/files in /etc/modules/6.1.x-meson64. The directory structures of kernel 6.1.11, 6.1.30 nd 6.1.42 are the same. However, kernel 6.1.11 seems to expect that the modules are compressed, it loads the modules and everything is fine for me. The filenames all end with .ko.xz while the modules from kernel 6.1.30 (Armbian bookworm version in the official download images) and the modules for kernel 6.1.42 are not compressed. Those filenames all end with .ko. I didn't check whether compressing the required modules (in kernel/drivers/firmware) can be loaded by the kernels 6.1.30 or 6.1.42 if they are compressed. So, if the first error message you see during reboot shows an error regarding "Failed: ...loading kernel modules" then try to compress the modules in /lib/modules/6.1.30-meson64/kernel/drivers/firmware and reboot. Hope this helps.

Hi Unfortunately no solution here, but "mee too" only. My USB keyboard works fine, that's it. In addition, I suddenly no longer have any Ethernet hardware onboard. This happened sometime during last night. However, I suspect the kernel version which is unmaintained now, again. I am still using bullseye and kernel 6.1.42 - no alternative kernel is available on the internal disk. Are you using the same kernel version ? Thank you for your answer. Edit: The last known kernel version which worked was 6.1.11 from a backup done with fsarchiver. Will try to fall back to this one and report results, here. Of course, I don't know which kernel version broke the Ethernet port und the USB disk support. There is an error message during boot stating that kernel modules can not be loaded. Perhaps this is related ? We'll see. New edit: Solved ! I replaced the broken kernel (6.1.42) setup by using the 6.1.11 kernel. I think the problem is with the modules of kernel 6.1.42 and that's why the external disks are not detected. The output from kernel 6.1.42 showed an error message saying that loading kernel modules failed. The same happened to me with kernel 6.1.11 because I forgot to add the modules directory to /lib/modules. As soon as I added /lib/modules/6.1.11-meson64 correctly my nextcloud now boots and works. I'll disable automatic updates now.

It does not just reboot automatically, but it crashes after a few hours of use. In addition it automaticall shuts down after a while. I am sick of opening the case, remove the eMMC, put it on the Micro-SD adapter, put it into the Micro-SD slot on my computer, edit the apache2 startup script, put the eMMC back on board and start the Odroid N2 again and repeat this several times a day. OK, I'll try to figure out myself how to stop using /dev/zram1 as /var/log.

Hi I had to upgrade my Odroid N2 from Armbian Buster to Armbian Bullseye to run newer versions of Nextcloudpi (24.x). So I did the upgrade by writing the Armbian 22.05 version to the eMMC and started Armbian. Works fine. Then I installed a fresh nextcloudpi (23.0.2) which is the same version I used on Armbian Buster. Works fine. I added the external disks (ncdata und ncbackup). After a restore of my nextcloudpi backup everything was well. But the next reboot did nothing, i.e. no output on the HDMI Interface, nothing. So, I did a power-off and looked at the eMMC via a Micro-SD adapter I have. Of course, no logs, nothing. Of course, there are no logs in /var/log because of the /dev/zram1 mounted at /var/log. Mounting /dev/zram1 on /var/log might be good if everything works, but if a reboot fails, you are host. I found a work-around: disable apache2 startup by editing the systemd unit file - I couldn't use systemctl because the system did not boot. After plugging in the external disks (nc-automount was enabled) and correcting the unit file, I could start apache2 and nextcloudpi is running well - until the next reboot. So, how do I disable the mount of /dev/zram1 to /var/log to find the real culprit ? Is it sufficient to delay the start of apache2 by adding ExecStartPre=/bin/sleep 30 ? Thanks for your advice. Kind regards Norbert