During working on #150837 I discovered that `google-oslogin` test
started failing, and so did some of my development machines. Turns out
it was because nscd doesn't start by default; rather it's wanted by
NSS lookup targets, which are not always fired up.
To quote from section on systemd.special(7) on `nss-user-lookup.target`:
> All services which provide parts of the user/group database should be
> ordered before this target, and pull it in.
Following this advice and comparing our unit to official `sssd.service`
unit (which is a similar service), we now pull NSS lookup targets from
the service, while starting it with `multi-user.target`.
This removes the `services.dbus.socketActivated` and
`services.xserver.startDbusSession` options. Instead the user D-Bus
session is always socket activated.
Since systemd 243, docs were already steering users towards using
`journal`:
eedaf7f322
systemd 246 will go one step further, it shows warnings for these units
during bootup, and will [automatically convert these occurences to
`journal`](f3dc6af20f):
> [ 6.955976] systemd[1]: /nix/store/hwyfgbwg804vmr92fxc1vkmqfq2k9s17-unit-display-manager.service/display-manager.service:27: Standard output type syslog is obsolete, automatically updating to journal. Please update│······················
your unit file, and consider removing the setting altogether.
So there's no point of keeping `syslog` here, and it's probably a better
idea to just not set it, due to:
> This setting defaults to the value set with DefaultStandardOutput= in
> systemd-system.conf(5), which defaults to journal.
This effectively disables nscd's built-in hosts cache, which turns out
to be erratic in some cases.
We only use nscd these days as a more ABI-neutral NSS dispatcher
mechanism.
Local caching should still be possible with local resolvers in
/etc/resolv.conf (via the `dns` NSS module), or without local resolvers
via systemd-networkd (via the `resolve` nss module)
We don't set enable-cache to no due to
https://github.com/NixOS/nixpkgs/pull/50316#discussion_r241035226.
This prevents duplication in cross-compiled nixos machines. The
bootstrapped glibc differs from the natively compiled one, so we get
two glibc’s in the closure. To reduce closure size, just use
stdenv.cc.libc where available.
Since https://github.com/NixOS/nixpkgs/pull/61321, local-fs.target is
part of sysinit.target again, meaning units without
DefaultDependencies=no will automatically depend on it, and the manual
set dependencies can be dropped.
NixOS usually needs nscd just to have a single place where
LD_LIBRARY_PATH can be set to include all NSS modules, but nscd is also
useful if some of the NSS modules need to read files which are only
accessible by root.
For example, nixos/modules/config/ldap.nix needs this when
users.ldap.enable = true;
users.ldap.daemon.enable = false;
and users.ldap.bind.passwordFile exists. In that case, the module
creates an /etc/ldap.conf which is only readable by root, but which the
NSS module needs to read in order to find out what LDAP server to
connect to and with what credentials.
If nscd is started as root and configured with the server-user option in
nscd.conf, then it gives each NSS module the opportunity to initialize
itself before dropping privileges. The initialization happens in the
glibc-internal __nss_disable_nscd function, which pre-loads all the
configured NSS modules for passwd, group, hosts, and services (but not
netgroup for some reason?) and, for each loaded module, calls an init
function if one is defined. After that finishes, nscd's main() calls
nscd_init() which ends by calling finish_drop_privileges().
There are provisions in systemd for using DynamicUser with a service
which needs to drop privileges itself, so this patch does that.
Thanks to @arianvp for pointing out that when DynamicUser is true,
systemd defaults the value of User to be the name of the unit, which in
this case is already "nscd".
These options were being set to the same value as the defaults that are
hardcoded in nscd. Delete them so it's clear which settings are actually
important for NixOS.
One exception is `threads 1`, which is different from the built-in
default of 4. However, both values are equivalent because nscd forces
the number of threads to be at least as many as the number of kinds of
databases it supports, which is 5.
nscd doesn't create any files outside of /run/nscd unless the nscd.conf
"persistent" option is used, which we don't do by default. Therefore it
doesn't matter what UID/GID we run this service as, so long as it isn't
shared with any other running processes.
/run/nscd does need to be owned by the same UID that the service is
running as, but systemd takes care of that for us thanks to the
RuntimeDirectory directive.
If someone wants to turn on the "persistent" option, they need to
manually configure users.users.nscd and systemd.tmpfiles.rules so that
/var/db/nscd is owned by the same user that nscd runs as.
In an all-defaults boot.isContainer configuration of NixOS, this removes
the only user which did not have a pre-assigned UID.
Previously this module created both /var/db/nscd and /run/nscd using
shell commands in a preStart script. Note that both of these paths are
hard-coded in the nscd source. (Well, the latter is actually
/var/run/nscd but /var/run is a symlink to /run so it works out the
same.)
/var/db/nscd is only used if the nscd.conf "persistent" option is turned
on for one or more databases, which it is not in our default config
file. I'm not even sure persistent mode can work under systemd, since
`nscd --shutdown` is not synchronous so systemd will always
unceremoniously kill nscd without reliably giving it time to mark the
databases as unused. Nonetheless, if someone wants to use that option,
they can ensure the directory exists using systemd.tmpfiles.rules.
systemd can create /run/nscd for us with the RuntimeDirectory directive,
with the added benefit of causing systemd to delete the directory on
service stop or restart. The default value of RuntimeDirectoryMode is
755, the same as the mode which this module was using before.
I don't think the `rm -f /run/nscd/nscd.pid` was necessary after NixOS
switched to systemd and used its PIDFile directive, because systemd
deletes the specified file after the service stops, and because the file
can't persist across reboots since /run is a tmpfs. Even if the file
still exists when nscd starts, it's only a problem if the pid it
contains has been reused by another process, which is unlikely. Anyway,
this change makes that deletion even less necessary, because now systemd
deletes the entire /run/nscd directory when the service stops.
This postStart step was introduced on 2014-04-24 with the comment that
"Nscd forks into the background before it's ready to accept
connections."
However, that was fixed upstream almost two months earlier, on
2014-03-03, with the comment that "This, along with setting the nscd
service type to forking in its systemd configuration file, allows
systemd to be certain that the nscd service is ready and is accepting
connections."
The fix was released several months later in glibc 2.20, which was
merged in NixOS sometime before 15.09, so it certainly should be safe to
remove this workaround by now.
The geoclue module now lets us set application config. This should make
it more robust in environments that don't provide a geoclue agent.
Fixes#44725.
