Eval does not stop on the first assertion failure it encouters.
Instead, it tries to evaluate all assertions and returns with a list of
those that failed.
This means our very top `config.mailserver.stateVersion != null`
assertion does not gate against any other assertions trying to compare
null against an integer.
The error prior to this commit can be reproduced by removing
`mailserver.stateVersion = 999;` in tests/lib/config.nix and then trying
to evaluate any of the tests:
~~~bash
# nix eval --raw .#checks.x86_64-linux.internal-unstable
error:
… while evaluating the attribute 'outPath'
at /nix/store/syvnmj3hhckkbncm94kfkbl76qsdqqj3-source/lib/customisation.nix:421:7:
420| drv.drvPath;
421| outPath =
| ^
422| assert condition;
… while calling the 'getAttr' builtin
at «internal»:1:500:
(stack trace truncated; use '--show-trace' to show the full trace)
error: cannot compare null with an integer
~~~
Per the dovecot documentation[0] we were previously running with an
unsupported home directory configuration, because we shared them among
all virtual users at /var/vmail.
After resolving this by creating per user home directories at
/var/vmail/%{domain}/%{user} this now also overlaps with the location of
the Maildir, which is not recommended.
As a result we now need to migrate our Maildirs into
/var/vmail/%{domain}/%{user}/mail, for which a small shell script is
provided as part of this change.
The script is included in the documentation because we cannot provide it
in time for users, because they might already be seeing the relevant
assertion and there is no safe waiting period that would allow us to skip
shipping it like that.
[0] https://doc.dovecot.org/2.3/configuration_manual/mail_location/
We have other tests that are minimal, e.g. the multiple test. And this
test wasn't even hooked up in flake.nix, so I'm doubtful that we really
need it.
These will never suceed while running the tests in the Nix sandbox, and
skipping them leads to very noticable (~51%) speedups.
Before:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 151.737 s ± 1.074 s [User: 0.310 s, System: 0.289 s]
Range (min … max): 150.321 s … 153.512 s 10 runs
```
After:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 74.010 s ± 0.746 s [User: 0.269 s, System: 0.266 s]
Range (min … max): 72.814 s … 75.190 s 10 runs
```
Provides a small (~7.5%) reduction in the test runtime measured for the external
test:
Before:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 151.737 s ± 1.074 s [User: 0.310 s, System: 0.289 s]
Range (min … max): 150.321 s … 153.512 s 10 runs
```
After:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 140.647 s ± 1.092 s [User: 0.331 s, System: 0.296 s]
Range (min … max): 138.536 s … 142.298 s 10 runs
```
Enabling the rspamd debug log drowns out everything else and should be
selected explicitly as needed.
The external test does not require it and removing it makes it much
(~40.5%) faster, since it now does not block on terminal output anymore.
Before:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 151.737 s ± 1.074 s [User: 0.310 s, System: 0.289 s]
Range (min … max): 150.321 s … 153.512 s 10 runs
```
After:
```
Benchmark 1: nix build .#hydraJobs.x86_64-linux.external-unstable --rebuild
Time (mean ± σ): 90.531 s ± 0.557 s [User: 0.054 s, System: 0.045 s]
Range (min … max): 89.579 s … 91.278 s 10 runs
```
I'm working on deprecating the top-level options, that configure main.cf
upstream in nixpkgs. With this change we stay ahead of the curve.
The `networks_style` option already defaults to `host` since Postfix 3.0,
so I dropped the setting.
```
$ postconf -d | grep networks_style
mynetworks_style = ${{$compatibility_level} <level {2} ? {subnet} : {host}}
````
- Groups settings between server and client
- Uses a range comparator for supported TLS versions
- Prune excluded primitives to what affects the supported TLS versions
It has been default enabled since Postfix 3.9 and can still be configured
from the NixOS option mentioned in the removal warning.
Removing the option makes our interface leaner.
Information is based on https://www.postfix.org/smtp-smuggling.html#long.
The sslCert and sslKey options are going away, because they do too much,
e.g. provision the keypair for client certificate authentication, which
is not at all what we want or need.
