There is a nasty race condition in the cups tests.
To understand what is going on, one must first note that
printers are installed in the vms with ensure-printers.service,
which is started as part of multi-user.target.
ensure-printers.service in turn triggers a start of
cups.service as it needs to connect to the local cups daemon.
This is what happens when the test runs:
1 the test waits for cups.socket or cups.service to start up
(subtest "Make sure that cups is up on both sides...")
2 after cups.service started
(it starts even in the "socket" case,
triggered by ensure-printers.service),
ensure-printers.service is started
3 the test tries to connect to the cups daemons via curl
(subtest "HTTP server is available too")
4 the test verifies the required printers are installed
("lpstat -a" called by subtest "LP status checks")
Usually, 3 needs some time, so ensure-printers.service
already installed all printers that are required by 4.
But if 3 is too fast, or if ensure-printers.service is too slow,
4 fails to find the printers it is looking for.
One can provoke the problem by adding
> systemd.services.ensure-printers.serviceConfig.ExecStartPre = "/run/current-system/sw/bin/sleep 10";
to the `nodes.client` configuration.
The commit at hand fixes the problem by changing 1:
Instead of waiting for cups,
it now waits for ensure-printers.service
(which in turn waits for cups.service and cups.socket).
This is also in accordance with the
subtest description in the code that promises to
"Make sure that cups is up [...] and printers are set up".
Prior to this contribution, every boot with a default configuration was
considered `ConditionFirstBoot=true` by systemd, since /etc/machine-id
was not commited to disk.
This also extends the systemd with a check for subsequent boots not
being considered first boots.
Adds explicit network configuration for etcd service
Waits for etcd to be fully healthy before running tests
Makes endpoint configuration explicit in etcdctl commands
Migrate to pkgs/by-name,
and update the test so that it passes for all versions
This version is added as EOL, since NetBox 4.1 is out,
but it might be still useful in case of an upgrade issue.
Modernize it. This allows the test to be extended, and pkgs to be
reused (later) to speed up evaluations a bit.
I believe this also makes it run on darwin hosts, but my linux-builder's
disk is too small to fit the massive closure of this test.
(cherry picked from commit 1396a03bee18a0993a4f3e97fda8938ff61c2918)
Modernize it. This allows the test to be extended, and pkgs to be
reused (later) to speed up evaluations a bit.
I believe this also makes it run on darwin hosts, but my linux-builder's
disk is too small to fit the massive closure of this test.
(cherry picked from commit 8c06d2cf667106dd440e7c140e70051dc1c321cb)
- remove dead code
- pass around a lot less redundant stuff
- add a timeout to the read so it can actually fail when characters are dropped
- run the input reader in systemd-cat so we can see the errors on console
This does not actually fix the flakiness in the tests, but it should make it
easier to find.
Before this change, the hash of the etc metadata image was included in
the mount unit that's responsible for mounting this metadata image in the
initrd.
And because this metadata image changes with every change to the etc
contents, the initrd would be rebuild every time as well.
This can lead to a lot of rebuilds (especially when revision info is
included in /etc/os-release) and all these initrd archives use up a lot of
space on the ESP.
With this change, we instead include a symlink to the metadata image in the
top-level directory, in the same way as we already do for things like init and
prepare-root, and we deduce the store path from the init= kernel parameter,
in the same way as we already do to find the path to init and prepare-root.
Doing so avoids rebuilding the initrd all the time.
The URL scheme for downloading plugins has changed a long time ago and
the used URL is dead. Gerrit only throws an error since it can't load
the plugin but it continues to boot. However, instead of maintaining
URLs to 3rdparty plugins, which end up dead anyway, just drop it. The
test should cover Gerrit and not 3rd party plugins.
Also, while on it, drop the setting `plugins.allowRemoteAdmin = true`
since it's not needed.
Signed-off-by: Felix Singer <felixsinger@posteo.net>
In order to emulate the `nixos-rebuild switch` that is called if the EC2
user data is a nix expression, run the switch-to-configuration script
for the current running config.
