This allows for the disko scripts to be ran in more foreign environments
where we can't assume the basic utilities coreutils provides are
available or aren't 1-for-1 compatible (e.g. busybox utilities).
I.e. instead of effectively ignoring `discardPolicy` when combined with
`randomEncryption`, or adding a second option requiring the user to
repeat themselves (DRY violation), just enable both when requested via
`discardPolicy`.
Without this `discardPolicy` is a NOP when `randomEncryption` is
enabled.
This hopefully fixes the large memory consumption we have seen in the
past. Also presumably coreutils cp is more tuned for performance than
whatever nix does.
closure-info has one flaw, it's possible that this file contains stale
information when build are not reproducible and the local checksum
doesn't match what the remote build created.
The work-around here is that we disregard the hashes and re-compute with
what nar's we actually have locally instead.
The destroy script already has a bash interpreter in use, but when the
`disk-deactivate` script is ran, it calls out to /usr/bin/env to query
for a bash interpreter. When running the destroyScript in unique places
(such as the stage-1 initrd), /usr/bin/env might not exist, so we can
make destroyScript more self-contained by reusing the same bash.
The option `disko.devices.lvm_vg.<name>.lvs.<name>.priority` depends on
a different option that's unset by default breaking the manual build
like this:
… while evaluating the attribute 'optionalValue.value'
at /nix/store/lwyjz70qh12nq6cb7fixl85vryzxqm3c-source/lib/modules.nix:856:5:
855|
856| optionalValue =
| ^
857| if isDefined then { value = mergedValue; }
… while evaluating a branch condition
at /nix/store/lwyjz70qh12nq6cb7fixl85vryzxqm3c-source/lib/modules.nix:857:7:
856| optionalValue =
857| if isDefined then { value = mergedValue; }
| ^
858| else {};
(stack trace truncated; use '--show-trace' to show the full trace)
error: The option `disko.devices.lvm_vg.<name>.lvs.<name>.size' is used but not defined.
This is especially noticeable when using `documentation.nixos.includeAllModules` locally.
Fixed the issue itself by adding a `defaultText` and added the
`disko-doc` job to the `checks` output so that CI hopefully catches this
from now on.
On 4Kn disks, 2048-sector alignment is equivalent to 8 MiB which is
excessive. Moreover, there is no point in specifying _any_ alignment
at all because sgdisk defaults to 1 MiB already.
Fixes#497.
The vmTools in nixpkgs shuts the VM down with poweroff -f, which does
not cleanly umount the filesystems. This may lead to filesystem
corruption on the newly created image.
Also refactor `zpool` type to use `zfs_fs` to construct the root dataset
for the zpool, which means we no longer need to duplicate the dataset
create and mount logic inside the `zpool` type.
If the zpool's root dataset is not the rootfs and gets mounted on
creation, the actual rootfs will get mounted later and shadow the
current mountpoint.
Running `zfs unmount` is the easiest way to unmount the zpool's root
dataset on creation without messing up the value of the `mountpoint`
setting.
