newedge/liminix
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

convert rst to asciidoc

Browse Source 
for i in *.rst ; do pandoc -f rst -t asciidoc -o `basename $i .rst`.adoc $i ;done
This commit is contained in:
Daniel Barlow
2025-04-06 18:16:53 +01:00
parent 936eb1a7cd
commit de861a2ee0
17 changed files with 1689 additions and 1783 deletions
Download Patch File Download Diff File Expand all files Collapse all files

470
doc/configuration.adoc Normal file
View File

@@ -0,0 +1,470 @@
== Configuration
There are many things you can specify in a configuration, but these are
the ones you most commonly need to change:
* which services (processes) to run
* what packages to install
* permitted users and groups
* Linux kernel configuration options
* Busybox applets
* filesystem layout
=== Modules
*Modules* are a means of abstraction which allow "bundling" of
configuration options related to a common purpose or theme. For example,
the `+dnsmasq+` module defines a template for a dnsmasq service, ensures
that the dnsmasq package is installed, and provides a dnsmasq user and
group for the service to run as. The `+ppp+` module defines a service
template and also enables various PPP-related kernel configuration.
Not all modules are included in the configuration by default, because
that would mean that the kernel (and the Busybox binary providing common
CLI tools) was compiled with many unnecessary bells and whistles and
therefore be bigger than needed. (This is not purely an academic concern
if your device has little flash storage). Therefore, specifying a
service is usually a two-step process. For example, to add an NTP
service you first add `+modules/ntp+` to your `+imports+` list, then you
create a service by calling `+config.system.service.ntp.build { .... }+`
with the appropriate service-dependent configuration parameters.
[source,nix]
----
let svc = config.system.service;
in {
# ...
imports = [
./modules/ntp
# ....
];
config.services.ntp = svc.ntp.build {
pools = { "pool.ntp.org" = ["iburst"]; };
makestep = { threshold = 1.0; limit = 3; };
};
----
Merely including the module won't define the service on its own: it only
creates the template in `+config.system.service.foo+` and you have to
create an actual service using the template. This is an intentional
choice to allow the creation of multiple differently-configured services
based on the same template - perhaps e.g. when you have multiple
networks (VPNs etc) in different trust domains, or you want to run two
SSH daemons on different ports. (For the background to this, please
refer to the `+architecture decision record <adr/module-system>+`)
[TIP]
====
Liminix modules should be quite familiar (but also different) if you
already know how to use NixOS modules. We use the NixOS module
infrastructure code, meaning that you should recognise the syntax, the
type system, the rules for combining configuration values from different
sources. We don't use the NixOS modules themselves, because the
underlying system is not similar enough for them to work.
====
[[configuration-services]]
=== Services
In Liminix a service is any kind of long-running task or process on the
system, that is managed (started, stopped, and monitored) by a service
supervisor. A typical SOHO router might have services to
* answer DHCP and DNS requests from the LAN
* provide a wireless access point
* connect using PPPoE or L2TP to an upstream network
* start/stop the firewall
* enable/disable IP packet forwarding
* mount filesystems
(Some of these might not be considered services using other definitions
of the term: for example, this L2TP process would be a "client" in the
client/server classification; and enabling packet forwarding doesn't
require any long-lived process - just a setting to be toggled. However,
there is value in being able to use the same abstractions for all the
things to manage them and specify their dependency relationships - so in
Liminix "everything is a service")
The service supervision system enables service health monitoring,
restart of unhealthy services, and failover to "backup" services when a
primary service fails or its dependencies are unavailable. The intention
is that you have a framework in which you can specify policy
requirements like "ethernet wan dhcp-client should be restarted if it
crashes, but if it can't start because the hardware link is down, then
4G ppp service should be started instead".
Any attribute in [.title-ref]#config.services# will become part of the
default set of services that s6-rc will try to bring up. Services are
usually started at boot time, but *controlled services* are those that
are required only in particular contexts. For example, a service to
mount a USB backup drive should run only when the drive is attached to
the system. Liminix currently implements three kinds of controlled
service:
* "uevent-rule" service controllers use sysfs/uevent to identify when
particular hardware devices are present, and start/stop a controlled
service appropriately.
* the "round-robin" service controller is used for service failover: it
allows you to specify a list of services and runs each of them in turn
until it exits, then runs the next.
* the "health-check" service wraps another service, and runs a "health
check" command at regular intervals. When the health check fails,
indicating that the wrapped service is not working, it is terminated and
allowed to restart.
==== Runtime secrets (external vault)
Secrets (such as wifi passphrases, PPP username/password, SSH keys, etc)
that you provide as literal values in `+configuration.nix+` are
processed into into config files and scripts at build time, and
eventually end up in various files in the (world-readable)
`+/nix/store+` before being baked into a flashable image. To change a
secret - whether due to a compromise, or just as part of to a routine
key rotation - you need to rebuild the configuration and potentially
reflash the affected devices.
To avoid this, you may instead use a "secrets service", which is a
mechanism for your device to fetch secrets from a source external to the
Nix store, and create at runtime the configuration files and scripts
that start the services which require them.
Not every possible parameter to every possible service is configurable
using a secrets service. Parameters which can be configured this way are
those with the type `+liminix.lib.types.replacable+`. At the time this
document was written, these include:
* ppp (pppoe and l2tp): `+username+`, `+password+`
* ssh: `+authorizedKeys+`
* hostapd: all parameters (most likely to be useful for
`+wpa_passphrase+`)
To use a runtime secret for any of these parameters:
* create a secrets service to specify the source of truth for secrets
* use the `+outputRef+` function in the service parameter to specify the
secrets service and path
For example, given you had an HTTPS server hosting a JSON file with the
structure
[source,json]
----
"ssh": {
"authorizedKeys": {
"root": [ "ssh-rsa ....", "ssh-rsa ....", ... ]
"guest": [ "ssh-rsa ....", "ssh-rsa ....", ... ]
}
}
----
you could use a `+configuration.nix+` fragment something like this to
make those keys visible to ssh:
[source,nix]
----
services.secrets = svc.secrets.outboard.build {
name = "secret-service";
url = "http://10.0.0.1/secrets.json";
username = "secrets";
password = "liminix";
interval = 30; # minutes
dependencies = [ config.services.lan ];
};
services.sshd = svc.ssh.build {
authorizedKeys = outputRef config.services.secrets "ssh/authorizedKeys";
};
----
There are presently two implementations of a secrets service:
===== Outboard secrets (HTTPS)
This service expects a URL to a JSON file containing all the secrets.
You may specify a username and password along with the URL, which are
used if the file is password-protected (HTTP Basic authentication). Note
that this is not a protection against a malicious local user: the
username and password are normal build-time parameters so will be
readable in the Nix store. This is a mitigation against the URL being
accidentally discovered due to e.g. a log file or error message on the
server leaking.
===== Tang secrets (encrypted local file)
Aternatively, secrets may be stored locally on the device, in a file
that has been encrypted using https://github.com/latchset/tang[Tang].
____
Tang is a server for binding data to network presence.
This sounds fancy, but the concept is simple. You have some data, but
you only want it to be available when the system containing the data is
on a certain, usually secure, network.
____
[source,nix]
----
services.secrets = svc.secrets.tang.build {
name = "secret-service";
path = "/run/mnt/usbstick/secrets.json.jwe";
interval = 30; # minutes
dependencies = [ config.services.mount-usbstick ];
};
----
The encryption uses the same scheme/algorithm as
https://github.com/latchset/clevis[Clevis] : you may use the
https://github.com/latchset/clevis?tab=readme-ov-file#pin-tang[Clevis
instructions] to encrypt the file on another host and then copy it to
your Liminix device, or you can use `+tangc encrypt+` to encrypt
directly on the device. (That latter approach may pose a chicken/egg
problem if the device needs secrets to boot up and run the services you
are relying on in order to login).
==== Writing services
For the most part, for common use cases, hopefully the services you need
will be defined by modules and you will only have to pass the right
parameters to `+build+`.
Should you need to create a custom service of your own devising, use the
[.title-ref]#oneshot# or [.title-ref]#longrun# functions:
* a "longrun" service is the "normal" service concept: it has a `+run+`
action which describes the process to start, and it watches that process
to restart it if it exits. The process should not attempt to daemonize
or "background" itself, otherwise s6-rc will think it died. Whatever it
prints to standard output/standard error will be logged.
[source,nix]
----
config.services.cowsayd = pkgs.liminix.services.longrun {
name = "cowsayd";
run = "${pkgs.cowsayd}/bin/cowsayd --port 3001 --breed hereford";
# don't start this until the lan interface is ready
dependencies = [ config.services.lan ];
}
----
* a "oneshot" service doesn't have a process attached. It consists of
`+up+` and `+down+` actions which are bits of shell script that are run
at the appropriate points in the service lifecycle
[source,nix]
----
config.services.greenled = pkgs.liminix.services.oneshot {
name = "greenled";
up = ''
echo 17 > /sys/class/gpio/export
echo out > /sys/class/gpio/gpio17/direction
echo 0 > /sys/class/gpio/gpio17/value
'';
down = ''
echo 0 > /sys/class/gpio/gpio17/value
'';
}
----
Services may have dependencies: as you see above in the `+cowsayd+`
example, it depends on some service called `+config.services.lan+`,
meaning that it won't be started until that other service is up.
==== Service outputs
Outputs are a mechanism by which a service can provide data which may be
required by other services. For example:
* the DHCP client service can expect to receive nameserver address
information as one of the fields in the response from the DHCP server:
we provide that as an output which a dependent service for a stub name
resolver can use to configure its upstream servers.
* a service that creates a new network interface (e.g. ppp) will provide
the name of the interface (`+ppp0+`, or `+ppp1+` or `+ppp7+`) as an
output so that a dependent service can reference it to set up a route,
or to configure firewall rules.
A service `+myservice+` should write its outputs as files in
`+/run/services/outputs/myservice+`: you can look around this directory
on a running Liminix system to see how it's used currently. Usually we
use the `+in_outputs+` shell function in the `+up+` or `+run+`
attributes of the service:
[source,shell]
----
(in_outputs ${name}
for i in lease mask ip router siaddr dns serverid subnet opt53 interface ; do
(printenv $i || true) > $i
done)
----
The outputs are just files, so technically you can read them using
anything that can read a file. Liminix has two "preferred" mechanisms,
though:
===== One-off lookups
In any context that ends up being evaluated by the shell, use `+output+`
to print the value of an output
[source,nix]
----
services.defaultroute4 = svc.network.route.build {
via = "$(output ${services.wan} address)";
target = "default";
dependencies = [ services.wan ];
};
----
===== Continuous updates
The downside of using shell functions in downstream service startup
scripts is that they only run when the service starts up: if a service
output _changes_, the downstream service would have to be restarted to
notice the change. Sometimes this is OK but other times the downstream
has no other need to restart, if it can only get its new data.
For this case, there is the `+anoia.svc+` Fennel library, which allows
you to write a simple loop which is iterated over whenever a service's
outputs change. This code is from
`+modules/dhcp6c/acquire-wan-address.fnl+`
[source,fennel]
----
(fn update-addresses [wan-device addresses new-addresses exec]
;; run some appropriate "ip address [add|remove]" commands
)
(fn run []
(let [[state-directory wan-device] arg
dir (svc.open state-directory)]
(accumulate [addresses []
v (dir:events)]
(update-addresses wan-device addresses
(or (v:output "address") []) system))))
----
The `+output+` method seen here accepts a filename (relative to the
service's output directory), or a directory name. It returns the first
line of that file, or for directories it returns a table (Lua's
key/value datastructure, similar to a hash/dictionary) of the outputs in
that directory.
===== Output design considerations
For preference, outputs should be short and simple, and not require
downstream services to do complicated parsing in order to use them.
Shell commands in Liminix are run using the Busybox shell which doesn't
have the niceties of an advanced shell like Bash let alone those of a
real programming language.
Note also that the Lua `+svc+` library only reads the first line of each
output.
=== Module implementation
Modules in Liminix conventionally live in
`+modules/somename/default.nix+`. If you want or need to write your own,
you may wish to refer to the examples there in conjunction with reading
this section.
A module is a function that accepts `+{lib, pkgs, config, ... }+` and
returns an attrset with keys `+imports, options config+`.
* `+imports+` is a list of paths to the other modules required by this
one
* `+options+` is a nested set of option declarations
* `+config+` is a nested set of option definitions
The NixOS manual section
https://nixos.org/manual/nixos/stable/#sec-writing-modules[Writing NixOS
Modules] is a quite comprehensive reference to writing NixOS modules,
which is also mostly applicable to Liminix except that it doesn't cover
service templates.
==== Service templates
To expose a service template in a module, it needs the following:
* an option declaration for `+system.service.myservicename+` with the
type of `+liminix.lib.types.serviceDefn+`
[source,nix]
----
options = {
system.service.cowsay = mkOption {
type = liminix.lib.types.serviceDefn;
};
};
----
* an option definition for the same key, which specifies where to import
the service template from (often `+./service.nix+`) and the types of its
parameters.
[source,nix]
----
config.system.service.cowsay = config.system.callService ./service.nix {
address = mkOption {
type = types.str;
default = "0.0.0.0";
description = "Listen on specified address";
example = "127.0.0.1";
};
port = mkOption {
type = types.port;
default = 22;
description = "Listen on specified TCP port";
};
breed = mkOption {
type = types.str;
default = "British Friesian"
description = "Breed of the cow";
};
};
----
Then you need to provide the service template itself, probably in
`+./service.nix+`:
[source,nix]
----
{
# any nixpkgs package can be named here
liminix
, cowsayd
, serviceFns
, lib
}:
# these are the parameters declared in the callService invocation
{ address, port, breed} :
let
inherit (liminix.services) longrun;
inherit (lib.strings) escapeShellArg;
in longrun {
name = "cowsayd";
run = "${cowsayd}/bin/cowsayd --address ${address} --port ${builtins.toString port} --breed ${escapeShellArg breed}";
}
----
[TIP]
====
Not relevant to module-based services specifically, but a common gotcha
when specifiying services is forgetting to transform "rich" parameter
values into text when composing a command for the shell to execute. Note
here that the port number, an integer, is stringified with `+toString+`,
and the name of the breed, which may contain spaces, is escaped with
`+escapeShellArg+`
====
==== Types
All of the NixOS module types are available in Liminix. These
Liminix-specific types also exist in `+pkgs.liminix.lib.types+`:
* `+service+`: an s6-rc service
* `+interface+`: an s6-rc service which specifies a network interface
* `+serviceDefn+`: a service "template" definition
In the future it is likely that we will extend this to include other
useful types in the networking domain: for example; IP address, network
prefix or netmask, protocol family and others as we find them.