| 7 Dec 2022 |
 Emil Thorsøe | stat /boot/loader/random-seed or similar | 14:25:35 |
| Emil Thorsøe | not sure, if this would be a thing that would work with stage1 systemd | 14:26:09 |
 ElvishJerricco | In theory that should be irrelevant. systemd-boot should be updating the random-seed | 14:28:52 |
| ElvishJerricco | In reply to @jkarlson:kapsi.fi
stat /boot/loader/random-seed or similar Looks like it was modified last time I booted. | 14:32:00 |
| Emil Thorsøe | thanks | 14:47:00 |
| Emil Thorsøe | do you use stage1 systemd? | 14:47:17 |
| ElvishJerricco | I do, but I am pretty confident that doesn't have anything to do with it | 14:51:28 |
| ElvishJerricco | btw the code in systemd-random-seed I was looking at is apparently new: https://github.com/systemd/systemd/pull/25319 | 14:52:02 |
| ElvishJerricco | So right now, systemd-boot is supposed to update the ESP random-seed if you have a "system token" efi variable (which I think is supposed to be setup by bootctl) | 14:52:46 |
| ElvishJerricco | And in the future, on shutdown, systemd-random-seed.service will cause the ESP random seed to be updated as well | 14:53:05 |
| ElvishJerricco | * And in a future version, on shutdown, systemd-random-seed.service will cause the ESP random seed to be updated as well | 14:53:24 |
|  threedee joined the room. | 20:18:03 |
| 8 Dec 2022 |
|  Sergei Lukianov joined the room. | 03:05:59 |
| Sergei Lukianov | hi, not sure if it's a best/right place to ask about it, but I've faced an issue with permissions that need to be granted to a systemd service to be able to run smartctl for nvme drives. I've specifically talking about prometheus-smartctl-exporter. It was recently fixed and work correctly for HDDs but it's not working for NVMe in my setup. Thank you in advance if there are some ideas!
https://github.com/NixOS/nixpkgs/blob/96d7f290878dec43e2688940467d24499c208cd9/nixos/modules/services/monitoring/prometheus/exporters/smartctl.nix#L42 | 03:10:50 |
|  ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 05:11:54 |
| ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 06:32:46 |
| ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 06:37:30 |
| ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 06:39:22 |
| Emil Thorsøe | In reply to @elvishjerricco:matrix.org
I do, but I am pretty confident that doesn't have anything to do with it maybe you are on unstable that maybe has new systemd? | 06:39:29 |
| ElvishJerricco | Emil Thorsøe: I am on a very recent unstable but it's still only on 251.7, not 252 yet | 06:40:47 |
| ElvishJerricco | * Emil Thorsøe: I am on a very recent nixos-unstable but it's still only on 251.7, not 252 yet | 06:41:26 |
| Emil Thorsøe | ack | 06:42:20 |
| Sergei Lukianov | Fixed by adding services.udev.extraRules = ''
SUBSYSTEM=="nvme", KERNEL=="nvme[0-9]*", GROUP="disk"
''; | 10:03:39 |
 hexa | I was wondering why systemd didn't provide that rule | 10:15:50 |
| hexa | so right now nvme0 is owned by root:root, but the namespaces aber owned by root:disk | 10:16:25 |
| ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 14:22:18 |
| ★.¸¸¸. •°´Nameless☆*´¨`*☽ changed their profile picture. | 14:28:15 |
| Sergei Lukianov | hexa yeah, I wasn't able to find any reasoning for it, but at least it works :) and doesn't look too bad | 18:42:24 |
| hexa | 😄 | 18:50:52 |
|  terrestrialpedestrian joined the room. | 19:09:11 |
