In 87a19e9048 I merged staging-next into master using the GitHub gui as intended.
In ac241fb7a5 I merged master into staging-next for the next staging cycle, however, I accidentally pushed it to master.
Thinking this may cause trouble, I reverted it in 0be87c7979. This was however wrong, as it "removed" master.
This reverts commit 0be87c7979.
I merged master into staging-next but accidentally pushed it to master.
This should get us back to 87a19e9048.
This reverts commit ac241fb7a5, reversing
changes made to 76a439239e.
Minor incompatibilities due to moving to upstream defaults:
- capabilities are used instead of systemd.socket units
- the control socket moved:
/run/kresd/control -> /run/knot-resolver/control/1
- cacheDir moved and isn't configurable anymore
- different user+group names, without static IDs
Thanks Mic92 for multiple ideas.
There is no need to stop/start the unit when the machine is online or
offline.
This should fix the shutdown locking issues.
nixos zerotier: sometimes it doesn't shutdown
Deperecates the interfaces option which was used to generate a host:port
list whereas the port was always hardcoded to 53. This unifies the
listen configuration for plain and TLS sockets and allows to specify a
port without an address for wildcard binds.
Systemd dependencies for scripted mode
were refactored according to analysis in #34586.
networking.vswitches can now be used with systemd-networkd,
although they are not supported by the daemon, a nixos receipe
creates the switch and attached required interfaces (just like
the scripted version).
Vlans and internal interfaces are implemented following the
template format i.e. each interface is
described using an attributeSet (vlan and type at the moment).
If vlan is present, then interface is added to the vswitch with
given tag (access mode). Type internal enabled vswitch to create
interfaces (see openvswitch docs).
Added configuration for configuring supported openFlow version on
the vswitch
This commit is a split from the original PR #35127.
A centralized list for these renames is not good because:
- It breaks disabledModules for modules that have a rename defined
- Adding/removing renames for a module means having to find them in the
central file
- Merge conflicts due to multiple people editing the central file
As part of the networking.* name space cleanup, connman should be moved
to services.connman. The same will happen for example with
networkmanager in a separate PR.
The new description should give more clear understanding of when to
edit the option.
I used NixOS to set up a DNS server that is authoritative for certain
zones. The description of the `cacheNetworks` option made me think I
needed to set it to `"any"` to allow people to query the zone I set
up. Reading the source of the module would have clarified my
understanding, but at the time I just read the description and thought
little of it. Later I discovered I was getting tons of DNS requests
and presumably being used for a DNS amplification attack or similar.
I have fixed the problem now, but I would like the option to have a
clearer description so others don't make the same mistake I did.
Add a virtual user system based around pam and a Berkeley
user database.
Adding the:
- localRoot
- userDbPath
- allowWriteableChroot
- virtualUseLocalPrivs
Vsftpd options.
Ever since setting up bonding the `wpa_supplicant-unit-start` script has
been failing. This is because the file `bonding_masters` in
`/sys/class/net/` is *not* a directory containing `uevent`.
Adding a test to verify the `uevent` path to be sourced exists resolves
the problem.
The two new options make it possible to create the interface in one namespace
and move it to a different one, as explained at https://www.wireguard.com/netns/.
In cases where you boot up really quickly (like in the VM test on a
non-busy host), tinydns might want to bind before the loopback interface
is fully up. Order tinydns after network.target to fix that.
Incorrect merging of modules resulted in dhcpcd being enabled causing flaky network connection.
https://github.com/NixOS/nixpkgs/pull/64364
Fixing it uncovered an infinite recursion from the same commit, previously masked by the incorrect merge.
We can just drop the `mkDefault` for `networking.wireless.enable` as it is already `false` by default.
Closes: https://github.com/NixOS/nixpkgs/issues/72416
This change ensures that `dhcpcd.service` is restarted as soon as the
exit hook changes. I use this hook to do additional configuration for my
network (like setting a route via the given gateway to my WireGuard) and
when changing parts of this exit hook I'd expect to get this activated
when switching to my new configuration.
This is a good example of a package/module that should be distributed
externally (e.g. as a flake [1]): it's not stable yet so anybody who
seriously wants to use it will want to use the upstream repo. Also,
it's highly specialized so NixOS is not really the right place at the
moment (every NixOS module slows down NixOS evaluation for everybody).
[1] https://github.com/edolstra/jormungandr/tree/flake
It seems that dnsdist doesn't actually request CAP_NET_BIND_SERVICE, which is why normally it's executed and root and setuids to another, unprivileged, user. This means that as it is, dnsdist will be unable to bind to any port under 1024 and will fail with access denied.
Removing CAP_SETGID and CAP_SETUID is also related to this as we don't actually change the uid or gid after the fact as we use DynamicUser. (That part isn't strictly NEEDED but there's no reason to have those capabilities if we don't use them).
There are also some additional sandboxing we can remove from the service definition as they are assumed true or strict by DynamicUser: specifically PrivateTmp and ProtectSystem respectively.
ProtectHome is still there, despite being assumed read-only as setting it to true means they are seen as empty. I don't think it really matters as I don't know if systemd will ignore it or not, but I didn't see any reason to go hunting for excuses to make it a bigger change.
This option was removed because allowing (multiple) regular users to
override host entries affecting the whole system opens up a huge attack
vector. There seem to be very rare cases where this might be useful.
Consider setting system-wide host entries using networking.hosts,
provide them via the DNS server in your network, or use
networking.networkmanager.appendNameservers to point your system to
another (local) nameserver to set those entries.
Otherwise connecting simply fails:
VPN connection: failed to connect: 'La création du fichier « /tmp/lib/NetworkManager-fortisslvpn/0507e3ef-f0e0-4153-af64-b3d9a025877c.config.XSB19Z » a échoué : No such file or directory'
This fixes a regression from bb649d96b0.
There were permission problems, when the preStart script tried to copy
the smokeping.fcgi file over the old file.
This introduces an option wifi.backend to the networkmanager module.
Co-authored-by: Cole Mickens <cole.mickens@gmail.com>
Co-authored-by: worldofpeace <worldofpeace@protonmail.ch>
This commits makes it clearer to a novice reader how to configure several
diferent types of SSID connections that were otherwise obscurely documented
Resolves#66650
These are the leftovers of an older PR.
a. Send messages to auditd if auditing is enabled.
b. Add missing dbus configuration if dnsmasq is used for DNS
Fixes problems such as:
systemd[1]: Failed to put bus name to hashmap: File exists
systemd[1]: dbus-org.freedesktop.nm-dispatcher.service: Two services allocated for the same bus name org.freedesktop.nm_dispatcher, refusing operation.
Problem is that systemd treats symlinks to files outside the service
path differently, causing our old workaround to look like two separate services.
These symlinks are intended to be a means for manually emulating
the behavior of the `Alias=` directive in these services.
Unfortunately even making these symlinks relative isn't enough,
since they don't make it to where it matters--
that only makes the links in /etc/static/systemd/system/*
relative, with systemd still being shown non-relative links
in /etc/systemd/system/*.
To fix this, drop all of this at the package level
and instead simply specify the aliases in the NixOS modules.
Also handle the same for modemmanager,
since the networkmanager NixOS module also handles that.
Fixes problems such as:
systemd[1]: Failed to put bus name to hashmap: File exists
systemd[1]: dbus-org.freedesktop.nm-dispatcher.service: Two services allocated for the same bus name org.freedesktop.nm_dispatcher, refusing operation.
Problem is that systemd treats symlinks to files outside the service
path differently, causing our old workaround to look like two separate services.
These symlinks are intended to be a means for manually emulating
the behavior of the `Alias=` directive in these services.
Unfortunately even making these symlinks relative isn't enough,
since they don't make it to where it matters--
that only makes the links in /etc/static/systemd/system/*
relative, with systemd still being shown non-relative links
in /etc/systemd/system/*.
To fix this, drop all of this at the package level
and instead simply specify the aliases in the NixOS modules.
Also handle the same for modemmanager,
since the networkmanager NixOS module also handles that.
This option was added in 6336048c58 but it
is essentially a complete duplicate of the existing cfg.servers and
there seems to be no reason to keep maintaining it.
Furthermore, it requires annoying duplication if you try to do option
merging, e.g. merging in sets into your configuration.nix that add
`services.chrony.initstepslew` options will overwrite the servers option
unless you keep it, but that means you just have to duplicate
config.networking.timeServers again anyway which is an implementation
detail!
Signed-off-by: Austin Seipp <aseipp@pobox.com>
'iburst' allows chrony to make very quick adjustments to the clock by
doing a couple rapid measurements outside of the default 'minpoll'
option. This helps improve rapid time adjustment at boot, and is enabled
by default.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
This is reckless, ill-advised, pointless, and I will be scorned for it,
but it makes me feel a lot better.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
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.
The `keys.target` is used to indicate whether all NixOps keys were
successfully uploaded on an unattended reboot. However this can cause
startup issues e.g. with NixOS containers (see #67265) and can block
boots even though this might not be needed (e.g. with a dovecot2
instance running that doesn't need any of the NixOps keys).
As described in the NixOps manual[1], dependencies to keys should be
defined like this now:
``` nix
{
systemd.services.myservice = {
after = [ "secret-key.service" ];
wants = [ "secret-key.service" ];
};
}
```
However I'd leave the issue open until it's discussed whether or not to
keep `keys.target` in `nixpkgs`.
[1] https://nixos.org/nixops/manual/#idm140737322342384
When NetworkManager is configured to not manage all interfaces, it's
perfectly fine to have the rest be managed by the standard nixos
wireless scripts.
I use
networking.networkmanager.unmanaged = [
"*" "except:type:wwan" "except:type:gsm"
];
to control everything using networking.wireless except for the mobile
LTE modem which only works with NetworkManager.
The following configuration generates a systemd unit that doesn't
start.
```nix
{
services.syncthing = {
enable = true;
user = "my-user";
};
}
```
It fails with
```
systemd[1]: Started Syncthing service.
systemd[6745]: syncthing.service: Failed to determine group credentials: No such process
systemd[6745]: syncthing.service: Failed at step GROUP spawning /nix/store/n1ydz3i08nqp1ajc50ycy1zribmphqc9-syncthing-1.1.4-bin/bin/syncthing: No such process
systemd[1]: syncthing.service: Main process exited, code=exited, status=216/GROUP
systemd[1]: syncthing.service: Failed with result 'exit-code'.
```
This is due to the fact that `syncthing` group (default) is not
created if the user is overridden.
Add a separate check for setting up the default group, so that
user/group are created independently.
This is a refactor of how resolvconf is managed on NixOS. We split it
into a separate service which is enabled internally depending on whether
we want /etc/resolv.conf to be managed by it. Various services now take
advantage of those configuration options.
We also now use systemd instead of activation scripts to update
resolv.conf.
NetworkManager now uses the right option for rc-manager DNS
automatically, so the configuration option shouldn't be exposed.
We were already creating a group for the user under which to run syncthing but
we were defaulting to running as `nogroup`.
Additionally, use `install` instead of multiple calls to mkdir/cp/chown.
Add an option to set the rc-manager parameter in NetworkManager.conf,
which controls how NetworkManager handles resolv.conf. This sets the
default rc-manager to "resolvconf", which solves #61490. It
additionally allows the user to change rc-manager without interference
from configuration activations.
Applies OpenWRT's noscan patch to hostapd and the relevant option to
the hostapd module.
This noscan patch adds a new `noscan` option allowing us to create
some overlapping BSSs in HT40+/- mode.
Note: this option is disabled by default, we leave this up to the end
user whether it should be enabled or not.
Not being able to create those overlapping BSSs is basically
preventing us to use 802.11n in any urban area where chances to
overlap with another SSID are extremely high.
The patch we are using is a courtesy of the openwrt team and is
applied to the defaul hostapd package in both OpenWRT and Archlinux.
The the extraConfig variable is added below the domain variable in the
ddclient config file. The domain variable should always be last.
(cherry picked from commit ba0ba6dc7934a6b4cc5d4090739a3a1c839afe67)
Previously each oneshot peer service only ran once and was not
restarted together with the interface unit. Because of this,
defined peers were missing after restarting their corresponding
interface unit.
Co-Authored-By: Franz Pletz <fpletz@fnordicwalking.de>
This can lead to unnecessary failures if the kernel module is already
loaded:
Jun 06 12:38:50 chef bglisn9bz0y5403vdw9hny0ij43r41jk-unit-script-wireguard-wg0-start[13261]: modprobe: FATAL: Module wireguard not found in directory /run/booted-system/kernel-modules/lib/modules/4.19.36
Types are now specified for all options.
The fixed uid and gid for the avahi user have been removed
and the user avahi is now in the group avahi.
The the generic opening of the firewall for UDP port 5353 is
now optional, but still defaults to true.
The option `extraServiceFiles` was added to specify avahi
service definitions, which are then placed in `/etc/avahi/services`.
Remove the btsync module. Bittorrent Sync was renamed to Resilio Sync in
2016, which is supported by the resilio module. Since Resilio Sync had
some security updates since 2016, it is not safe to run Bittorrent Sync
anymore.
Before, changing any peers caused the entire WireGuard interface to
be torn down and rebuilt. By configuring each peer in a separate
service we're able to only restart the affected peers.
Adding each peer individually also means individual peer
configurations can fail, but the overall interface and all other peers
will still be added.
A WireGuard peer's internal identifier is its public key. This means
it is the only reliable identifier to use for the systemd service.
When calling reload, bird attempts to reload the file that was given in
the command line. As the change of ${configFile} is never picked up,
bird will just reload the old file.
This way, the configuration is placed at a known location and updated.
As a oneshot service, if the startup failed it would never be attempted again.
This is problematic when peer's addresses require DNS. DNS may not be reliably available at
the time wireguard starts. Converting this to a simple service with Restart
and RestartAfter directives allows the service to be reattempted, but at
the cost of losing the oneshot semantics.
Signed-off-by: Maximilian Bosch <maximilian@mbosch.me>
Passwords should not be stored in plain text by default. On existing
installations the next time a users user accounts will automatically
be upgraded from plain to hashed one-by-one as they log in.
With `sshd -t` config validation for SSH is possible. Until now, the
config generated by Nix was applied without any validation (which is
especially a problem for advanced config like `Match` blocks).
When deploying broken ssh config with nixops to a remote machine it gets
even harder to fix the problem due to the broken ssh that makes reverts
with nixops impossible.
This change performs the validation in a Nix build environment by
creating a store path with the config and generating a mocked host key
which seems to be needed for the validation. With a broken config, the
deployment already fails during the build of the derivation.
The original attempt was done in #56345 by adding a submodule for Match
groups to make it harder screwing that up, however that made the module
far more complex and config should be described in an easier way as
described in NixOS/rfcs#42.
I noticed xinetd process doesn't get exec'd on launch, exec here so the bash
process doesn't stick around.
Signed-off-by: William Casarin <jb55@jb55.com>
The type of ZNC's config option specifies that a configuration like
config.User.paul = null;
should be valid, which is useful for clearing/disabling property sets
like Users and Networks. However until now the config generator
implementation didn't actually cover null values, meaning you'd get an
error like
error: value is null while a set was expected, at /foo.nix:29:10
This fixes the implementation to correcly allow clearing of property
sets.
This allows configuring IP addresses on a tinc interface using
networking.interfaces."tinc.${n}".ipv[46].addresses.
Previously, this would fail with timeouts, because of the dependency
chain
tinc.${netname}.service
--after--> network.target
--after--> network-addresses-tinc.${n}.service (and network-link-…)
--after--> sys-subsystem-net-devices-tinc.${n}.device
But the network interface doesn't exist until tinc creates it! So
systemd waits in vain for the interface to appear, and by then the
network-addresses-* and network-link-* units have failed. This leads
to the network link not being brought up and the network addresses not
being assigned, which in turn stops tinc from actually working.
Ideally, private keys never leave the host they're generated on - like
SSH. Setting generatePrivateKeyFile to true causes the PK to be
generate automatically.
This is an implementation of wireguard support using wg-quick config
generation.
This seems preferrable to the existing wireguard support because
it handles many more routing and resolvconf edge cases than the
current wireguard support.
It also includes work-arounds to make key files work.
This has one quirk:
We need to set reverse path checking in the firewall to false because
it interferes with the way wg-quick sets up its routing.
Get these from upstream tox-node package instead.
This is likely to cause less maintenance overhead over time and
following upstream bootstrap node changes is automated.
Same problem as described in acbadcdbba.
When using multiple interfaces for wifi with `networking.wlanInterfaces`
and the interface for `hostapd` contains a dash, this will fail as
systemd escapes dashes in its device names.
The manpage claims that the "limit" in the setting::
<name>:[<limit>:]<regex>
is optional and defaults to zero, implying no limit.
However, tests confirmed that it actually isn't optional.
Without limit, the setting ``any:.*`` places
outbound jobs on infinite hold if no particular
modem was specified on the sendfax command line.
The new default value ``any:0:.*`` from
this commit uses any available modem to
send jobs if not modem was given to sendfax.
Before this change, only passwords not containing shell metacharacters could be
used, and because the password was passed as a command-line argument, local
users could (in a very small window of time) record the password and (in an
indefinity window of time) record the length of the password.
We also use the opportunity to add a call to `exec` in the systemd start
script, so that no shell needs to hang around waiting for iodine to stop.
Compatibility with other distributions/software and expectation
of users coming from other systems should have higher priority over consistency.
In particular this fixes#51375, where the NetworkManager-wait-online.service
broke as a result of this.
Systemd provides some functionality to escape strings that are supposed
to be part of a unit name[1]. This seems to be used for interface names
in `sys-subsystem-net-devices-{interface}.device` and breaks
wpa_supplicant if the wireless interface name has a dash which is
encoded to \x2d.
Such an interface name is rather rare, but used i.e. when configuring
multiple wireless interfaces with `networking.wlanInterfaces`[2] to have on
interface for `wpa_supplicant` and another one for `hostapd`.
[1] https://www.freedesktop.org/software/systemd/man/systemd-escape.html
[2] https://nixos.org/nixos/options.html#networking.wlaninterfaces
Add an ExecReload command to the prosody service, to allow reloading
prosody by sending SIGHUP to the main process, for example to update
certificates without restarting the server. This is exactly how the
`prosodyctl` tool does it.
Note: Currently there is a bug which prevents mod_http from reloading the
certificates properly: https://issues.prosody.im/1216.
Without this line, attempting to copy and paste non-ASCII characters
will result in error messages like the following (and pasting from the
server to the client will not work):
```
CLIPBOARD clipboard_send_data_response_for_text: 823 : ERROR: clipboard_send_data_response_for_text: bad string
```
While at it (see previous commit), using attrNames in combination with
length is a bit verbose for checking whether the filtered attribute set
is empty, so let's just compare it against an empty attribute set.
Signed-off-by: aszlig <aszlig@nix.build>
When generating values for the services.nsd.zones attribute using values
from pkgs, we'll run into an infinite recursion because the nsd module
has a condition on the top-level definition of nixpkgs.config.
While it would work to push the definition a few levels down, it will
still only work if we don't use bind tools for generating zones.
As far as I could see, Python support for BIND seems to be only needed
for the dnssec-* tools, so instead of using nixpkgs.config, we now
directly override pkgs.bind instead of globally in nixpkgs.
To illustrate the problem with a small test case, instantiating the
following Nix expression from the nixpkgs source root will cause the
mentioned infinite recursion:
(import ./nixos {
configuration = { lib, pkgs, ... }: {
services.nsd.enable = true;
services.nsd.zones = import (pkgs.writeText "foo.nix" ''
{ "foo.".data = "xyz";
"foo.".dnssec = true;
}
'');
};
}).vm
With this change, generating zones via import-from-derivation is now
possible again.
Signed-off-by: aszlig <aszlig@nix.build>
Cc: @pngwjpgh
This flag causes the shairport-sync server to attempt to daemonize, but it looks like systemd is already handling that. With the `-d` argument, shairport-sync immediately exits—it seems that something (systemd I'm guessing?) is sending it SIGINT or SIGTERM.
The [upstream systemd unit](https://github.com/mikebrady/shairport-sync/blob/master/scripts/shairport-sync.service.in#L10) doesn't pass `-d`.
As the comment notes, restarts/exits of dhcpcd generally require
restarting the NTP service since, if name resolution fails for a pool of
servers, the service might break itself. To be on the safe side, try
restarting Chrony in these instances, too.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
Setting the server list to be empty is useful e.g. for hardware-only
or virtualized reference clocks that are passed through to the system
directly. In this case, initstepslew has no effect, so don't emit it.
Signed-off-by: Austin Seipp <aseipp@pobox.com>