| 18 Aug 2025 |
Sylvie (They/She) 🏳️⚧️ | I'll just hope that this also won't cause trouble when I deploy updates | 12:28:37 |
lassulus | you could check this:
If $SYSTEMD_RELAX_ESP_CHECKS=1 is set the validation checks for
the ESP are relaxed, and the path specified with --esp-path= may
refer to any kind of file system on any kind of partition.
Similarly, $SYSTEMD_RELAX_XBOOTLDR_CHECKS=1 turns off some
validation checks for the Extended Boot Loader partition.
| 12:29:07 |
lassulus | not sure they apply here, seems like the wrong way around | 12:29:37 |
lassulus | but it should work I think | 12:30:12 |
Sylvie (They/She) 🏳️⚧️ | hmmm yeah | 12:37:05 |
Sylvie (They/She) 🏳️⚧️ | I tried just passing that in via --impure and it did the trick | 12:37:14 |
Sylvie (They/She) 🏳️⚧️ | now how/where do I add that to my config 😅 | 12:37:37 |
lassulus | maybe add a nixos-install wrapper | 12:37:56 |
lassulus | environment.systemPackages = [
(pkgs.writeScriptBin "nixos-install" ''
export SYSTEMD_RELAX_ESP_CHECKS=1
exec {pkgs.nixos-install}/bin/nixos-install "$@"
'')
]
| 12:39:10 |
lassulus | or something like that | 12:39:13 |
Sylvie (They/She) 🏳️⚧️ | hmmmm | 12:47:39 |
Sylvie (They/She) 🏳️⚧️ | https://github.com/systemd/systemd/issues/12468#issuecomment-698782077
And yes, sd-boot writes to the ESP if boot counting/boot assessment is enabled.
| 12:49:43 |
Sylvie (They/She) 🏳️⚧️ | I'd like to turn that off lol | 12:49:48 |
Sylvie (They/She) 🏳️⚧️ | I don't see it in the loader.conf options | 12:50:09 |
Sylvie (They/She) 🏳️⚧️ | it sounds like setting that up is a rather involved process fwiw | 12:53:28 |
Sylvie (They/She) 🏳️⚧️ | I'd say I can worry about that once I actually have two disks 😅 | 12:53:44 |
Sylvie (They/She) 🏳️⚧️ | hmmmm | 12:59:29 |
Sylvie (They/She) 🏳️⚧️ |  Download photo.jpg | 12:59:47 |
Sylvie (They/She) 🏳️⚧️ | there is no partition table there, but the mountpoints disko generates seem to assume there is? | 13:00:30 |
Sylvie (They/She) 🏳️⚧️ | nix eval .#nixosConfigurations.boop.config.fileSystems."/boot" | nix run nixpkgs#nixfmt-rfc-style
{
autoFormat = false;
autoResize = false;
depends = [ ];
device = "/dev/md/esp";
enable = true;
encrypted = {
blkDev = null;
enable = false;
keyFile = null;
label = null;
};
formatOptions = null;
fsType = "vfat";
label = null;
mountPoint = "/boot";
neededForBoot = false;
noCheck = false;
options = [ "umask=0077" ];
overlay = {
lowerdir = null;
upperdir = null;
useStage1BaseDirectories = true;
workdir = null;
};
stratis = {
poolUuid = null;
};
}
| 13:01:42 |
Sylvie (They/She) 🏳️⚧️ | this looks fine to me though 🤔 | 13:01:51 |
Sylvie (They/She) 🏳️⚧️ | seems mdmonitor.service failed cause my config is invalid | 13:12:25 |
Sylvie (They/She) 🏳️⚧️ | dunno if that's related hmmm | 13:12:49 |
Sylvie (They/She) 🏳️⚧️ | is that service required for md to function? | 13:13:04 |
Sylvie (They/She) 🏳️⚧️ | the name doesn't make it sound like it lol | 13:13:11 |
lassulus | usually not | 13:13:15 |
lassulus | but the device should be mountable | 13:13:31 |
Sylvie (They/She) 🏳️⚧️ | hmmm where do I find the units for mounts | 13:20:02 |
momo | * Disko can be downloaded normally, but I don't understand why the subsequent operations are performed directly in the ISO instead of after creating partitions with Disko. My host has 8GB RAM and a 60GB storage, so logically, whether using tmpfs or the hard drive, resources shouldn't be exhausted. | 13:20:06 |
momo | * Disko can be downloaded normally, but I don't understand why the subsequent operations are performed directly in the ISO instead of after creating partitions with Disko. My host has 8GB RAM and 60GB storage, so logically, whether using tmpfs or the hard drive, resources shouldn't be exhausted. | 13:20:18 |