On Mon, 16 Jan 2023 at 15:45, Markus Schaaf <markuschaaf@gmail.com> wrote:
for a long time now, I'm seeing unclean shutdowns of SATA drives on different machines running Arch. There is a nice write-up about the problem here: https://lore.kernel.org/lkml/20170410232118.GA4816@khazad-dum.debian.net/ If you are willing to google with the right keywords, you will find a lot of similar reports all over the interwebs. I can remember a configuration setting for the Linux kernel, which existed in ancient times, to delay the final step of powering off the machine for some time. Is there anything like that in the distributed kernels right now? Any knob I could turn?
you can probably make a shutdown service that delays the shutdown or even invokes some commands to make sure HDDs are fine? another way is to make a custom mkinitcpio hook that adds a shutdown script (alas I can't find the proper documentation about it) -- damjan