By some miracle, before, it was possible to reconnect to the `node1` without
doing any relevant dance.
But now we are direct booting (¿), it seems like we need to do the right things.
This introduces a `check_output` flag for `execute` because we do not want to steal the
messages from the backdoor service as we might execute the kexec too fast compared
to when we will reconnect.
Therefore, we will let the message in the pipe if needed.
There's no reason to use a bootloader when testing kexec, this is a feature
that reboots *directly* in the kernel, if anything, we should just direct boot the
kernel and reboots in the kernel.
A bootloader test really makes sense to test "default" systemctl kexec behavior which is already broken
because systemctl kexec will read the ESP to determine what to kexec by default.
This change removes the bespoke logic around identifying block devices.
Instead of trying to find the right device by iterating over
`qemu.drives` and guessing the right partition number (e.g.
/dev/vda{1,2}), devices are now identified by persistent names provided
by udev in /dev/disk/by-*.
Before this change, the root device was formatted on demand in the
initrd. However, this makes it impossible to use filesystem identifiers
to identify devices. Now, the formatting step is performed before the VM
is started. Because some tests, however, rely on this behaviour, a
utility function to replace this behaviour in added in
/nixos/tests/common/auto-format-root-device.nix.
Devices that contain neither a partition table nor a filesystem are
identified by their hardware serial number which is injecetd via QEMU
(and is thus persistent and predictable). PCI paths are not a reliably
way to identify devices because their availability and numbering depends
on the QEMU machine type.
This change makes the module more robust against changes in QEMU and the
kernel (non-persistent device naming) and by decoupling abstractions
(i.e. rootDevice, bootPartition, and bootLoaderDevice) enables further
improvement down the line.
It's supposed to be `memcache.distributed`, not an associative PHP array
named `memcache` with a key `distributed`.
This was probably never caught because the initial `grep -q` check in
the test was invalid: `redis-cli` prints nothing if no keys can be found
when not writing to a tty apparently.
a zfs fileSystems entry with an absolute (e.g. device) path rather than
a zfs dataser is parsed as an empty pool name, causing a doomed-to-fail
import job to be created as a boot dependency. Catch this as an assertion
* remove `with`
* replace specific hooks with attrsOf lines
To be flexible, should they change.
* make hooks with writeShellApplication
- Previously hooks would not build if they used a heredoc with `EOF`
- To shellcheck hooks
* format with nixpkgs-fmt
* remove removed option module
Lemmy checks the environment variable before the configuration file;
i.e. if the file is used to configure the database but the environment
variable is set to anything, the connection will fail because it'll
ignore the file. This was the previous behavior.
Now, the environment variable will be unset unless the user explicitly
chooses to set it, which makes the file-based configuration function
correctly. It's also possible to manually set the environment variable,
which has the major advantage of working around [this issue][0], which
prevents certain setups from working.
[0]: https://github.com/LemmyNet/lemmy/issues/2945
This option conditionally adds the `CAP_NET_RAW` capability to the service,
which is mandatory for enabling the integrated DHCP server.
It also adds another test case to validate that the DHCP server successfully
provides IP addresses to clients.
Co-authored-by: Sandro <sandro.jaeckel@gmail.com>
