I. Installation
1. Obtaining NixOS
2. Installing NixOS
2.1. UEFI Installation
2.2. Booting from a USB Drive
3. Changing the Configuration
4. Upgrading NixOS
4.1. Automatic Upgrades
II. Configuration
5. Configuration Syntax
5.1. NixOS Configuration File
5.2. Abstractions
5.3. Modularity
5.4. Syntax Summary
6. Package Management
6.1. Declarative Package Management
6.1.1. Customising Packages
6.1.2. Adding Custom Packages
6.2. Ad-Hoc Package Management
7. User Management
8. File Systems
8.1. LUKS-Encrypted File Systems
9. X Window System
10. Networking
10.1. NetworkManager
10.2. Secure Shell Access
10.3. IPv4 Configuration
10.4. IPv6 Configuration
10.5. Firewall
10.6. Wireless Networks
10.7. Ad-Hoc Configuration
11. Linux Kernel
12. PostgreSQL
12.1. Configuring
12.2. Upgrading
12.3. Options
III. Administration
13. Service Management
14. Rebooting and Shutting Down
15. User Sessions
16. Control Groups
17. Logging
18. Cleaning the Nix Store
19. Container Management
19.1. Imperative Container Management
19.2. Declarative Container Specification
19.3. Container Networking
20. Troubleshooting
20.1. Boot Problems
20.2. Maintenance Mode
20.3. Rolling Back Configuration Changes
20.4. Nix Store Corruption
20.5. Network Problems
IV. Development
21. Getting the Sources
22. Writing NixOS Modules
22.1. Option Declarations
22.2. Option Definitions
23. Building Specific Parts of NixOS
24. Building Your Own NixOS CD
25. NixOS Tests
25.1. Writing Tests
25.2. Running Tests
25.3. Running Tests interactively
26. Testing the Installer
A. Configuration Options
B. Release Notes
B.1. Release 15.09 (“Dingo”, 2015/09/30)
B.2. Release 14.12 (“Caterpillar”, 2014/12/30)
B.3. Release 14.04 (“Baboon”, 2014/04/30)
B.4. Release 13.10 (“Aardvark”, 2013/10/31)

This manual describes how to install, use and extend NixOS, a Linux distribution based on the purely functional package management system Nix.

If you encounter problems, please report them on the mailing list or on the #nixos channel on Freenode. Bugs should be reported in NixOS’ GitHub issue tracker.

This section describes how to obtain, install, and configure NixOS for first-time use.

Chapter 1. Obtaining NixOS

NixOS ISO images can be downloaded from the NixOS download page. There are a number of installation options. If you happen to have an optical drive and a spare CD, burning the image to CD and booting from that is probably the easiest option. Most people will need to prepare a USB stick to boot from. Unetbootin is recommended and the process is described in brief below. Note that systems which use UEFI require some additional manual steps. If you run into difficulty a number of alternative methods are presented in the NixOS Wiki.

As an alternative to installing NixOS yourself, you can get a running NixOS system through several other means:

  • Using virtual appliances in Open Virtualization Format (OVF) that can be imported into VirtualBox. These are available from the NixOS download page.

  • Using AMIs for Amazon’s EC2. To find one for your region and instance type, please refer to the list of most recent AMIs.

  • Using NixOps, the NixOS-based cloud deployment tool, which allows you to provision VirtualBox and EC2 NixOS instances from declarative specifications. Check out the NixOps homepage for details.

Chapter 2. Installing NixOS

  1. Boot from the CD.

  2. The CD contains a basic NixOS installation. (It also contains Memtest86+, useful if you want to test new hardware). When it’s finished booting, it should have detected most of your hardware.

  3. The NixOS manual is available on virtual console 8 (press Alt+F8 to access).

  4. Login as root and the empty password.

  5. If you downloaded the graphical ISO image, you can run start display-manager to start KDE.

  6. The boot process should have brought up networking (check ip a). Networking is necessary for the installer, since it will download lots of stuff (such as source tarballs or Nixpkgs channel binaries). It’s best if you have a DHCP server on your network. Otherwise configure networking manually using ifconfig.

    To manually configure the network on the graphical installer, first disable network-manager with systemctl stop network-manager.

  7. The NixOS installer doesn’t do any partitioning or formatting yet, so you need to that yourself. Use the following commands:

    • For partitioning: fdisk.

    • For initialising Ext4 partitions: mkfs.ext4. It is recommended that you assign a unique symbolic label to the file system using the option -L label, since this makes the file system configuration independent from device changes. For example:

      $ mkfs.ext4 -L nixos /dev/sda1

    • For creating swap partitions: mkswap. Again it’s recommended to assign a label to the swap partition: -L label.

    • For creating LVM volumes, the LVM commands, e.g.,

      $ pvcreate /dev/sda1 /dev/sdb1
      $ vgcreate MyVolGroup /dev/sda1 /dev/sdb1
      $ lvcreate --size 2G --name bigdisk MyVolGroup
      $ lvcreate --size 1G --name smalldisk MyVolGroup

    • For creating software RAID devices, use mdadm.

  8. Mount the target file system on which NixOS should be installed on /mnt, e.g.

    $ mount /dev/disk/by-label/nixos /mnt

  9. If your machine has a limited amount of memory, you may want to activate swap devices now (swapon device). The installer (or rather, the build actions that it may spawn) may need quite a bit of RAM, depending on your configuration.

  10. You now need to create a file /mnt/etc/nixos/configuration.nix that specifies the intended configuration of the system. This is because NixOS has a declarative configuration model: you create or edit a description of the desired configuration of your system, and then NixOS takes care of making it happen. The syntax of the NixOS configuration file is described in Chapter 5, Configuration Syntax, while a list of available configuration options appears in Appendix A, Configuration Options. A minimal example is shown in Example 2.2, “NixOS Configuration”.

    The command nixos-generate-config can generate an initial configuration file for you:

    $ nixos-generate-config --root /mnt

    You should then edit /mnt/etc/nixos/configuration.nix to suit your needs:

    $ nano /mnt/etc/nixos/configuration.nix

    If you’re using the graphical ISO image, other editors may be available (such as vim). If you have network access, you can also install other editors — for instance, you can install Emacs by running nix-env -i emacs.

    You must set the option boot.loader.grub.device to specify on which disk the GRUB boot loader is to be installed. Without it, NixOS cannot boot.

    Another critical option is fileSystems, specifying the file systems that need to be mounted by NixOS. However, you typically don’t need to set it yourself, because nixos-generate-config sets it automatically in /mnt/etc/nixos/hardware-configuration.nix from your currently mounted file systems. (The configuration file hardware-configuration.nix is included from configuration.nix and will be overwritten by future invocations of nixos-generate-config; thus, you generally should not modify it.)

    Note: Depending on your hardware configuration or type of file system, you may need to set the option boot.initrd.kernelModules to include the kernel modules that are necessary for mounting the root file system, otherwise the installed system will not be able to boot. (If this happens, boot from the CD again, mount the target file system on /mnt, fix /mnt/etc/nixos/configuration.nix and rerun nixos-install.) In most cases, nixos-generate-config will figure out the required modules.

    Examples of real-world NixOS configuration files can be found at

  11. Do the installation:

    $ nixos-install

    Cross fingers. If this fails due to a temporary problem (such as a network issue while downloading binaries from the NixOS binary cache), you can just re-run nixos-install. Otherwise, fix your configuration.nix and then re-run nixos-install.

    As the last step, nixos-install will ask you to set the password for the root user, e.g.

    setting root password...
    Enter new UNIX password: ***
    Retype new UNIX password: ***

  12. If everything went well:

    $ reboot

  13. You should now be able to boot into the installed NixOS. The GRUB boot menu shows a list of available configurations (initially just one). Every time you change the NixOS configuration (see Changing Configuration ), a new item is added to the menu. This allows you to easily roll back to a previous configuration if something goes wrong.

    You should log in and change the root password with passwd.

    You’ll probably want to create some user accounts as well, which can be done with useradd:

    $ useradd -c 'Eelco Dolstra' -m eelco
    $ passwd eelco

    You may also want to install some software. For instance,

    $ nix-env -qa \*

    shows what packages are available, and

    $ nix-env -i w3m

    install the w3m browser.

To summarise, Example 2.1, “Commands for Installing NixOS on /dev/sda shows a typical sequence of commands for installing NixOS on an empty hard drive (here /dev/sda). Example 2.2, “NixOS Configuration” shows a corresponding configuration Nix expression.

Example 2.1. Commands for Installing NixOS on /dev/sda

$ fdisk /dev/sda # (or whatever device you want to install on)
$ mkfs.ext4 -L nixos /dev/sda1
$ mkswap -L swap /dev/sda2
$ swapon /dev/sda2
$ mount /dev/disk/by-label/nixos /mnt
$ nixos-generate-config --root /mnt
$ nano /mnt/etc/nixos/configuration.nix
$ nixos-install
$ reboot

Example 2.2. NixOS Configuration

{ config, pkgs, ... }:

  imports =
    [ # Include the results of the hardware scan.

  boot.loader.grub.device = "/dev/sda";

  # Note: setting fileSystems is generally not
  # necessary, since nixos-generate-config figures them out
  # automatically in hardware-configuration.nix.
  #fileSystems."/".device = "/dev/disk/by-label/nixos";

  # Enable the OpenSSH server.
  services.sshd.enable = true;

2.1. UEFI Installation

NixOS can also be installed on UEFI systems. The procedure is by and large the same as a BIOS installation, with the following changes:

  • You should boot the live CD in UEFI mode (consult your specific hardware's documentation for instructions). You may find the rEFInd boot manager useful.

  • Instead of fdisk, you should use gdisk to partition your disks. You will need to have a separate partition for /boot with partition code EF00, and it should be formatted as a vfat filesystem.

  • You must set boot.loader.gummiboot.enable to true. nixos-generate-config should do this automatically for new configurations when booted in UEFI mode.

  • After having mounted your installation partition to /mnt, you must mount the boot partition to /mnt/boot.

  • You may want to look at the options starting with boot.loader.efi and boot.loader.gummiboot as well.

2.2. Booting from a USB Drive

For systems without CD drive, the NixOS live CD can be booted from a USB stick. For non-UEFI installations, unetbootin will work. For UEFI installations, you should mount the ISO, copy its contents verbatim to your drive, then either:

  • Change the label of the disk partition to the label of the ISO (visible with the blkid command), or

  • Edit loader/entries/nixos-livecd.conf on the drive and change the root= field in the options line to point to your drive (see the documentation on root= in the kernel documentation for more details).

Chapter 3. Changing the Configuration

The file /etc/nixos/configuration.nix contains the current configuration of your machine. Whenever you’ve changed something to that file, you should do

$ nixos-rebuild switch

to build the new configuration, make it the default configuration for booting, and try to realise the configuration in the running system (e.g., by restarting system services).

Warning: These commands must be executed as root, so you should either run them from a root shell or by prefixing them with sudo -i.

You can also do

$ nixos-rebuild test

to build the configuration and switch the running system to it, but without making it the boot default. So if (say) the configuration locks up your machine, you can just reboot to get back to a working configuration.

There is also

$ nixos-rebuild boot

to build the configuration and make it the boot default, but not switch to it now (so it will only take effect after the next reboot).

You can make your configuration show up in a different submenu of the GRUB 2 boot screen by giving it a different profile name, e.g.

$ nixos-rebuild switch -p test 

which causes the new configuration (and previous ones created using -p test) to show up in the GRUB submenu “NixOS - Profile 'test'”. This can be useful to separate test configurations from “stable” configurations.

Finally, you can do

$ nixos-rebuild build

to build the configuration but nothing more. This is useful to see whether everything compiles cleanly.

If you have a machine that supports hardware virtualisation, you can also test the new configuration in a sandbox by building and running a QEMU virtual machine that contains the desired configuration. Just do

$ nixos-rebuild build-vm
$ ./result/bin/run-*-vm

The VM does not have any data from your host system, so your existing user accounts and home directories will not be available. You can forward ports on the host to the guest. For instance, the following will forward host port 2222 to guest port 22 (SSH):

$ QEMU_NET_OPTS="hostfwd=tcp::2222-:22" ./result/bin/run-*-vm

allowing you to log in via SSH (assuming you have set the appropriate passwords or SSH authorized keys):

$ ssh -p 2222 localhost

Chapter 4. Upgrading NixOS

The best way to keep your NixOS installation up to date is to use one of the NixOS channels. A channel is a Nix mechanism for distributing Nix expressions and associated binaries. The NixOS channels are updated automatically from NixOS’s Git repository after certain tests have passed and all packages have been built. These channels are:

  • Stable channels, such as nixos-14.12. These only get conservative bug fixes and package upgrades. For instance, a channel update may cause the Linux kernel on your system to be upgraded from 3.4.66 to 3.4.67 (a minor bug fix), but not from 3.4.x to 3.11.x (a major change that has the potential to break things). Stable channels are generally maintained until the next stable branch is created.

  • The unstable channel, nixos-unstable. This corresponds to NixOS’s main development branch, and may thus see radical changes between channel updates. It’s not recommended for production systems.

  • Small channels, such as nixos-14.12-small or nixos-unstable-small. These are identical to the stable and unstable channels described above, except that they contain fewer binary packages. This means they get updated faster than the regular channels (for instance, when a critical security patch is committed to NixOS’s source tree), but may require more packages to be built from source than usual. They’re mostly intended for server environments and as such contain few GUI applications.

To see what channels are available, go to (Note that the URIs of the various channels redirect to a directory that contains the channel’s latest version and includes ISO images and VirtualBox appliances.)

When you first install NixOS, you’re automatically subscribed to the NixOS channel that corresponds to your installation source. For instance, if you installed from a 14.12 ISO, you will be subscribed to the nixos-14.12 channel. To see which NixOS channel you’re subscribed to, run the following as root:

$ nix-channel --list | grep nixos

To switch to a different NixOS channel, do

$ nix-channel --add nixos

(Be sure to include the nixos parameter at the end.) For instance, to use the NixOS 14.12 stable channel:

$ nix-channel --add nixos

If you have a server, you may want to use the “small” channel instead:

$ nix-channel --add nixos

And if you want to live on the bleeding edge:

$ nix-channel --add nixos

You can then upgrade NixOS to the latest version in your chosen channel by running

$ nixos-rebuild switch --upgrade

which is equivalent to the more verbose nix-channel --update nixos; nixos-rebuild switch.

Warning: It is generally safe to switch back and forth between channels. The only exception is that a newer NixOS may also have a newer Nix version, which may involve an upgrade of Nix’s database schema. This cannot be undone easily, so in that case you will not be able to go back to your original channel.

4.1. Automatic Upgrades

You can keep a NixOS system up-to-date automatically by adding the following to configuration.nix:

system.autoUpgrade.enable = true;

This enables a periodically executed systemd service named nixos-upgrade.service. It runs nixos-rebuild switch --upgrade to upgrade NixOS to the latest version in the current channel. (To see when the service runs, see systemctl list-timers.) You can also specify a channel explicitly, e.g. =;

This chapter describes how to configure various aspects of a NixOS machine through the configuration file /etc/nixos/configuration.nix. As described in Chapter 3, Changing the Configuration, changes to this file only take effect after you run nixos-rebuild.

Chapter 5. Configuration Syntax

The NixOS configuration file /etc/nixos/configuration.nix is actually a Nix expression, which is the Nix package manager’s purely functional language for describing how to build packages and configurations. This means you have all the expressive power of that language at your disposal, including the ability to abstract over common patterns, which is very useful when managing complex systems. The syntax and semantics of the Nix language are fully described in the Nix manual, but here we give a short overview of the most important constructs useful in NixOS configuration files.

5.1. NixOS Configuration File

The NixOS configuration file generally looks like this:

{ config, pkgs, ... }:

{ option definitions

The first line ({ config, pkgs, ... }:) denotes that this is actually a function that takes at least the two arguments config and pkgs. (These are explained later.) The function returns a set of option definitions ({ ... }). These definitions have the form name = value, where name is the name of an option and value is its value. For example,

{ config, pkgs, ... }:

{ services.httpd.enable = true;
  services.httpd.adminAddr = "";
  services.httpd.documentRoot = "/webroot";

defines a configuration with three option definitions that together enable the Apache HTTP Server with /webroot as the document root.

Sets can be nested, and in fact dots in option names are shorthand for defining a set containing another set. For instance, services.httpd.enable defines a set named services that contains a set named httpd, which in turn contains an option definition named enable with value true. This means that the example above can also be written as:

{ config, pkgs, ... }:

{ services = {
    httpd = {
      enable = true;
      adminAddr = "";
      documentRoot = "/webroot";

which may be more convenient if you have lots of option definitions that share the same prefix (such as services.httpd).

NixOS checks your option definitions for correctness. For instance, if you try to define an option that doesn’t exist (that is, doesn’t have a corresponding option declaration), nixos-rebuild will give an error like:

The option `services.httpd.enable' defined in `/etc/nixos/configuration.nix' does not exist.

Likewise, values in option definitions must have a correct type. For instance, services.httpd.enable must be a Boolean (true or false). Trying to give it a value of another type, such as a string, will cause an error:

The option value `services.httpd.enable' in `/etc/nixos/configuration.nix' is not a boolean.

Options have various types of values. The most important are:


Strings are enclosed in double quotes, e.g.

networking.hostName = "dexter";

Special characters can be escaped by prefixing them with a backslash (e.g. \").

Multi-line strings can be enclosed in double single quotes, e.g.

networking.extraHosts =
  '' other-localhost server

The main difference is that preceding whitespace is automatically stripped from each line, and that characters like " and \ are not special (making it more convenient for including things like shell code).


These can be true or false, e.g.

networking.firewall.enable = true;
networking.firewall.allowPing = false;


For example,

boot.kernel.sysctl."net.ipv4.tcp_keepalive_time" = 60;

(Note that here the attribute name net.ipv4.tcp_keepalive_time is enclosed in quotes to prevent it from being interpreted as a set named net containing a set named ipv4, and so on. This is because it’s not a NixOS option but the literal name of a Linux kernel setting.)


Sets were introduced above. They are name/value pairs enclosed in braces, as in the option definition

fileSystems."/boot" =
  { device = "/dev/sda1";
    fsType = "ext4";
    options = "rw,data=ordered,relatime";


The important thing to note about lists is that list elements are separated by whitespace, like this:

boot.kernelModules = [ "fuse" "kvm-intel" "coretemp" ];

List elements can be any other type, e.g. sets:

swapDevices = [ { device = "/dev/disk/by-label/swap"; } ];


Usually, the packages you need are already part of the Nix Packages collection, which is a set that can be accessed through the function argument pkgs. Typical uses:

environment.systemPackages =
  [ pkgs.thunderbird

postgresql.package = pkgs.postgresql90;

The latter option definition changes the default PostgreSQL package used by NixOS’s PostgreSQL service to 9.0. For more information on packages, including how to add new ones, see Section 6.1.2, “Adding Custom Packages”.

5.2. Abstractions

If you find yourself repeating yourself over and over, it’s time to abstract. Take, for instance, this Apache HTTP Server configuration:

  services.httpd.virtualHosts =
    [ { hostName = "";
        documentRoot = "/webroot";
        adminAddr = "";
        enableUserDir = true;
      { hostName = "";
        documentRoot = "/webroot";
        adminAddr = "";
        enableUserDir = true;
        enableSSL = true;
        sslServerCert = "/root/ssl-example-org.crt";
        sslServerKey = "/root/ssl-example-org.key";

It defines two virtual hosts with nearly identical configuration; the only difference is that the second one has SSL enabled. To prevent this duplication, we can use a let:

  exampleOrgCommon =
    { hostName = "";
      documentRoot = "/webroot";
      adminAddr = "";
      enableUserDir = true;
  services.httpd.virtualHosts =
    [ exampleOrgCommon
      (exampleOrgCommon // {
        enableSSL = true;
        sslServerCert = "/root/ssl-example-org.crt";
        sslServerKey = "/root/ssl-example-org.key";

The let exampleOrgCommon = ... defines a variable named exampleOrgCommon. The // operator merges two attribute sets, so the configuration of the second virtual host is the set exampleOrgCommon extended with the SSL options.

You can write a let wherever an expression is allowed. Thus, you also could have written:

  services.httpd.virtualHosts =
    let exampleOrgCommon = ...; in
    [ exampleOrgCommon
      (exampleOrgCommon // { ... })

but not { let exampleOrgCommon = ...; in ...; } since attributes (as opposed to attribute values) are not expressions.

Functions provide another method of abstraction. For instance, suppose that we want to generate lots of different virtual hosts, all with identical configuration except for the host name. This can be done as follows:

  services.httpd.virtualHosts =
      makeVirtualHost = name:
        { hostName = name;
          documentRoot = "/webroot";
          adminAddr = "";
      [ (makeVirtualHost "")
        (makeVirtualHost "")
        (makeVirtualHost "")
        (makeVirtualHost "")

Here, makeVirtualHost is a function that takes a single argument name and returns the configuration for a virtual host. That function is then called for several names to produce the list of virtual host configurations.

We can further improve on this by using the function map, which applies another function to every element in a list:

  services.httpd.virtualHosts =
      makeVirtualHost = ...;
    in map makeVirtualHost
      [ "" "" "" "" ];

(The function map is called a higher-order function because it takes another function as an argument.)

What if you need more than one argument, for instance, if we want to use a different documentRoot for each virtual host? Then we can make makeVirtualHost a function that takes a set as its argument, like this:

  services.httpd.virtualHosts =
      makeVirtualHost = { name, root }:
        { hostName = name;
          documentRoot = root;
          adminAddr = "";
    in map makeVirtualHost
      [ { name = ""; root = "/sites/"; }
        { name = ""; root = "/sites/"; }
        { name = ""; root = "/sites/"; }
        { name = ""; root = "/sites/"; }

But in this case (where every root is a subdirectory of /sites named after the virtual host), it would have been shorter to define makeVirtualHost as

makeVirtualHost = name:
  { hostName = name;
    documentRoot = "/sites/${name}";
    adminAddr = "";

Here, the construct ${...} allows the result of an expression to be spliced into a string.

5.3. Modularity

The NixOS configuration mechanism is modular. If your configuration.nix becomes too big, you can split it into multiple files. Likewise, if you have multiple NixOS configurations (e.g. for different computers) with some commonality, you can move the common configuration into a shared file.

Modules have exactly the same syntax as configuration.nix. In fact, configuration.nix is itself a module. You can use other modules by including them from configuration.nix, e.g.:

{ config, pkgs, ... }:

{ imports = [ ./vpn.nix ./kde.nix ];
  services.httpd.enable = true;
  environment.systemPackages = [ pkgs.emacs ];

Here, we include two modules from the same directory, vpn.nix and kde.nix. The latter might look like this:

{ config, pkgs, ... }:

{ services.xserver.enable = true;
  services.xserver.displayManager.kdm.enable = true;
  services.xserver.desktopManager.kde4.enable = true;
  environment.systemPackages = [ pkgs.kde4.kscreensaver ];

Note that both configuration.nix and kde.nix define the option environment.systemPackages. When multiple modules define an option, NixOS will try to merge the definitions. In the case of environment.systemPackages, that’s easy: the lists of packages can simply be concatenated. The value in configuration.nix is merged last, so for list-type options, it will appear at the end of the merged list. If you want it to appear first, you can use mkBefore:

boot.kernelModules = mkBefore [ "kvm-intel" ];

This causes the kvm-intel kernel module to be loaded before any other kernel modules.

For other types of options, a merge may not be possible. For instance, if two modules define services.httpd.adminAddr, nixos-rebuild will give an error:

The unique option `services.httpd.adminAddr' is defined multiple times, in `/etc/nixos/httpd.nix' and `/etc/nixos/configuration.nix'.

When that happens, it’s possible to force one definition take precedence over the others:

services.httpd.adminAddr = pkgs.lib.mkForce "";

When using multiple modules, you may need to access configuration values defined in other modules. This is what the config function argument is for: it contains the complete, merged system configuration. That is, config is the result of combining the configurations returned by every module[1]. For example, here is a module that adds some packages to environment.systemPackages only if services.xserver.enable is set to true somewhere else:

{ config, pkgs, ... }:

{ environment.systemPackages =
    if then
      [ pkgs.firefox
      [ ];

With multiple modules, it may not be obvious what the final value of a configuration option is. The command nixos-option allows you to find out:

$ nixos-option services.xserver.enable

$ nixos-option boot.kernelModules
[ "tun" "ipv6" "loop" ... ]

Interactive exploration of the configuration is possible using nix-repl, a read-eval-print loop for Nix expressions. It’s not installed by default; run nix-env -i nix-repl to get it. A typical use:

$ nix-repl '<nixos>'

nix-repl> config.networking.hostName

nix-repl> map (x: x.hostName)
[ "" "" ]

5.4. Syntax Summary

Below is a summary of the most important syntactic constructs in the Nix expression language. It’s not complete. In particular, there are many other built-in functions. See the Nix manual for the rest.

Basic values
"Hello world"A string
"${pkgs.bash}/bin/sh"A string containing an expression (expands to "/nix/store/hash-bash-version/bin/sh")
true, falseBooleans
123An integer
./foo.pngA path (relative to the containing Nix expression)
Compound values
{ x = 1; y = 2; }An set with attributes names x and y
{ = 1; }A nested set, equivalent to { foo = { bar = 1; }; }
rec { x = "foo"; y = x + "bar"; }A recursive set, equivalent to { x = "foo"; y = "foobar"; }
[ "foo" "bar" ]A list with two elements
"foo" + "bar"String concatenation
1 + 2Integer addition
"foo" == "f" + "oo"Equality test (evaluates to true)
"foo" != "bar"Inequality test (evaluates to true)
!trueBoolean negation
{ x = 1; y = 2; }.xAttribute selection (evaluates to 1)
{ x = 1; y = 2; }.z or 3Attribute selection with default (evaluates to 3)
{ x = 1; y = 2; } // { z = 3; }Merge two sets (attributes in the right-hand set taking precedence)
Control structures
if 1 + 1 == 2 then "yes!" else "no!"Conditional expression
assert 1 + 1 == 2; "yes!"Assertion check (evaluates to "yes!")
let x = "foo"; y = "bar"; in x + yVariable definition
with pkgs.lib; head [ 1 2 3 ]Add all attributes from the given set to the scope (evaluates to 1)
Functions (lambdas)
x: x + 1A function that expects an integer and returns it increased by 1
(x: x + 1) 100A function call (evaluates to 101)
let inc = x: x + 1; in inc (inc (inc 100))A function bound to a variable and subsequently called by name (evaluates to 103)
{ x, y }: x + yA function that expects a set with required attributes x and y and concatenates them
{ x, y ? "bar" }: x + yA function that expects a set with required attribute x and optional y, using "bar" as default value for y
{ x, y, ... }: x + yA function that expects a set with required attributes x and y and ignores any other attributes
{ x, y } @ args: x + yA function that expects a set with required attributes x and y, and binds the whole set to args
Built-in functions
import ./foo.nixLoad and return Nix expression in given file
map (x: x + x) [ 1 2 3 ]Apply a function to every element of a list (evaluates to [ 2 4 6 ])

[1] If you’re wondering how it’s possible that the (indirect) result of a function is passed as an input to that same function: that’s because Nix is a “lazy” language — it only computes values when they are needed. This works as long as no individual configuration value depends on itself.

Chapter 6. Package Management

This section describes how to add additional packages to your system. NixOS has two distinct styles of package management:

  • Declarative, where you declare what packages you want in your configuration.nix. Every time you run nixos-rebuild, NixOS will ensure that you get a consistent set of binaries corresponding to your specification.

  • Ad hoc, where you install, upgrade and uninstall packages via the nix-env command. This style allows mixing packages from different Nixpkgs versions. It’s the only choice for non-root users.

6.1. Declarative Package Management

With declarative package management, you specify which packages you want on your system by setting the option environment.systemPackages. For instance, adding the following line to configuration.nix enables the Mozilla Thunderbird email application:

environment.systemPackages = [ pkgs.thunderbird ];

The effect of this specification is that the Thunderbird package from Nixpkgs will be built or downloaded as part of the system when you run nixos-rebuild switch.

You can get a list of the available packages as follows:

$ nix-env -qaP '*' --description
nixos.firefox   firefox-23.0   Mozilla Firefox - the browser, reloaded

The first column in the output is the attribute name, such as nixos.thunderbird. (The nixos prefix allows distinguishing between different channels that you might have.)

To “uninstall” a package, simply remove it from environment.systemPackages and run nixos-rebuild switch.

6.1.1. Customising Packages

Some packages in Nixpkgs have options to enable or disable optional functionality or change other aspects of the package. For instance, the Firefox wrapper package (which provides Firefox with a set of plugins such as the Adobe Flash player) has an option to enable the Google Talk plugin. It can be set in configuration.nix as follows: nixpkgs.config.firefox.enableGoogleTalkPlugin = true;

Warning: Unfortunately, Nixpkgs currently lacks a way to query available configuration options.

Apart from high-level options, it’s possible to tweak a package in almost arbitrary ways, such as changing or disabling dependencies of a package. For instance, the Emacs package in Nixpkgs by default has a dependency on GTK+ 2. If you want to build it against GTK+ 3, you can specify that as follows:

environment.systemPackages = [ (pkgs.emacs.override { gtk = pkgs.gtk3; }) ];

The function override performs the call to the Nix function that produces Emacs, with the original arguments amended by the set of arguments specified by you. So here the function argument gtk gets the value pkgs.gtk3, causing Emacs to depend on GTK+ 3. (The parentheses are necessary because in Nix, function application binds more weakly than list construction, so without them, environment.systemPackages would be a list with two elements.)

Even greater customisation is possible using the function overrideDerivation. While the override mechanism above overrides the arguments of a package function, overrideDerivation allows changing the result of the function. This permits changing any aspect of the package, such as the source code. For instance, if you want to override the source code of Emacs, you can say:

environment.systemPackages =
  [ (pkgs.lib.overrideDerivation pkgs.emacs (attrs: {
      name = "emacs-25.0-pre";
      src = /path/to/my/emacs/tree;

Here, overrideDerivation takes the Nix derivation specified by pkgs.emacs and produces a new derivation in which the original’s name and src attribute have been replaced by the given values. The original attributes are accessible via attrs.

The overrides shown above are not global. They do not affect the original package; other packages in Nixpkgs continue to depend on the original rather than the customised package. This means that if another package in your system depends on the original package, you end up with two instances of the package. If you want to have everything depend on your customised instance, you can apply a global override as follows:

nixpkgs.config.packageOverrides = pkgs:
  { emacs = pkgs.emacs.override { gtk = pkgs.gtk3; };

The effect of this definition is essentially equivalent to modifying the emacs attribute in the Nixpkgs source tree. Any package in Nixpkgs that depends on emacs will be passed your customised instance. (However, the value pkgs.emacs in nixpkgs.config.packageOverrides refers to the original rather than overridden instance, to prevent an infinite recursion.)

6.1.2. Adding Custom Packages

It’s possible that a package you need is not available in NixOS. In that case, you can do two things. First, you can clone the Nixpkgs repository, add the package to your clone, and (optionally) submit a patch or pull request to have it accepted into the main Nixpkgs repository. This is described in detail in the Nixpkgs manual. In short, you clone Nixpkgs:

$ git clone git://
$ cd nixpkgs

Then you write and test the package as described in the Nixpkgs manual. Finally, you add it to environment.systemPackages, e.g.

environment.systemPackages = [ ];

and you run nixos-rebuild, specifying your own Nixpkgs tree:

$ nixos-rebuild switch -I nixpkgs=/path/to/my/nixpkgs

The second possibility is to add the package outside of the Nixpkgs tree. For instance, here is how you specify a build of the GNU Hello package directly in configuration.nix:

environment.systemPackages =
    my-hello = with pkgs; stdenv.mkDerivation rec {
      name = "hello-2.8";
      src = fetchurl {
        url = "mirror://gnu/hello/${name}.tar.gz";
        sha256 = "0wqd8sjmxfskrflaxywc7gqw7sfawrfvdxd9skxawzfgyy0pzdz6";
  [ my-hello ];

Of course, you can also move the definition of my-hello into a separate Nix expression, e.g.

environment.systemPackages = [ (import ./my-hello.nix) ];

where my-hello.nix contains:

with import <nixpkgs> {}; # bring all of Nixpkgs into scope

stdenv.mkDerivation rec {
  name = "hello-2.8";
  src = fetchurl {
    url = "mirror://gnu/hello/${name}.tar.gz";
    sha256 = "0wqd8sjmxfskrflaxywc7gqw7sfawrfvdxd9skxawzfgyy0pzdz6";

This allows testing the package easily:

$ nix-build my-hello.nix
$ ./result/bin/hello
Hello, world!

6.2. Ad-Hoc Package Management

With the command nix-env, you can install and uninstall packages from the command line. For instance, to install Mozilla Thunderbird:

$ nix-env -iA nixos.thunderbird

If you invoke this as root, the package is installed in the Nix profile /nix/var/nix/profiles/default and visible to all users of the system; otherwise, the package ends up in /nix/var/nix/profiles/per-user/username/profile and is not visible to other users. The -A flag specifies the package by its attribute name; without it, the package is installed by matching against its package name (e.g. thunderbird). The latter is slower because it requires matching against all available Nix packages, and is ambiguous if there are multiple matching packages.

Packages come from the NixOS channel. You typically upgrade a package by updating to the latest version of the NixOS channel:

$ nix-channel --update nixos

and then running nix-env -i again. Other packages in the profile are not affected; this is the crucial difference with the declarative style of package management, where running nixos-rebuild switch causes all packages to be updated to their current versions in the NixOS channel. You can however upgrade all packages for which there is a newer version by doing:

$ nix-env -u '*'

A package can be uninstalled using the -e flag:

$ nix-env -e thunderbird

Finally, you can roll back an undesirable nix-env action:

$ nix-env --rollback

nix-env has many more flags. For details, see the nix-env(1) manpage or the Nix manual.

Chapter 7. User Management

NixOS supports both declarative and imperative styles of user management. In the declarative style, users are specified in configuration.nix. For instance, the following states that a user account named alice shall exist:

users.extraUsers.alice =
  { isNormalUser = true;
    home = "/home/alice";
    description = "Alice Foobar";
    extraGroups = [ "wheel" "networkmanager" ];
    openssh.authorizedKeys.keys = [ "ssh-dss AAAAB3Nza... alice@foobar" ];

Note that alice is a member of the wheel and networkmanager groups, which allows her to use sudo to execute commands as root and to configure the network, respectively. Also note the SSH public key that allows remote logins with the corresponding private key. Users created in this way do not have a password by default, so they cannot log in via mechanisms that require a password. However, you can use the passwd program to set a password, which is retained across invocations of nixos-rebuild.

If you set users.mutableUsers to false, then the contents of /etc/passwd and /etc/group will be congruent to your NixOS configuration. For instance, if you remove a user from users.extraUsers and run nixos-rebuild, the user account will cease to exist. Also, imperative commands for managing users and groups, such as useradd, are no longer available.

A user ID (uid) is assigned automatically. You can also specify a uid manually by adding

    uid = 1000;

to the user specification.

Groups can be specified similarly. The following states that a group named students shall exist:

users.extraGroups.students.gid = 1000;

As with users, the group ID (gid) is optional and will be assigned automatically if it’s missing.

In the imperative style, users and groups are managed by commands such as useradd, groupmod and so on. For instance, to create a user account named alice:

$ useradd -m alice

The flag -m causes the creation of a home directory for the new user, which is generally what you want. The user does not have an initial password and therefore cannot log in. A password can be set using the passwd utility:

$ passwd alice
Enter new UNIX password: ***
Retype new UNIX password: ***

A user can be deleted using userdel:

$ userdel -r alice

The flag -r deletes the user’s home directory. Accounts can be modified using usermod. Unix groups can be managed using groupadd, groupmod and groupdel.

Chapter 8. File Systems

You can define file systems using the fileSystems configuration option. For instance, the following definition causes NixOS to mount the Ext4 file system on device /dev/disk/by-label/data onto the mount point /data:

fileSystems."/data" =
  { device = "/dev/disk/by-label/data";
    fsType = "ext4";

Mount points are created automatically if they don’t already exist. For device, it’s best to use the topology-independent device aliases in /dev/disk/by-label and /dev/disk/by-uuid, as these don’t change if the topology changes (e.g. if a disk is moved to another IDE controller).

You can usually omit the file system type (fsType), since mount can usually detect the type and load the necessary kernel module automatically. However, if the file system is needed at early boot (in the initial ramdisk) and is not ext2, ext3 or ext4, then it’s best to specify fsType to ensure that the kernel module is available.

8.1. LUKS-Encrypted File Systems

NixOS supports file systems that are encrypted using LUKS (Linux Unified Key Setup). For example, here is how you create an encrypted Ext4 file system on the device /dev/sda2:

$ cryptsetup luksFormat /dev/sda2

This will overwrite data on /dev/sda2 irrevocably.

Are you sure? (Type uppercase yes): YES
Enter LUKS passphrase: ***
Verify passphrase: ***

$ cryptsetup luksOpen /dev/sda2 crypted
Enter passphrase for /dev/sda2: ***

$ mkfs.ext4 /dev/mapper/crypted

To ensure that this file system is automatically mounted at boot time as /, add the following to configuration.nix:

boot.initrd.luks.devices = [ { device = "/dev/sda2"; name = "crypted"; } ];
fileSystems."/".device = "/dev/mapper/crypted";

Chapter 9. X Window System

The X Window System (X11) provides the basis of NixOS’ graphical user interface. It can be enabled as follows:

services.xserver.enable = true;

The X server will automatically detect and use the appropriate video driver from a set of drivers (such as vesa and intel). You can also specify a driver manually, e.g.

services.xserver.videoDrivers = [ "r128" ];

to enable’s xf86-video-r128 driver.

You also need to enable at least one desktop or window manager. Otherwise, you can only log into a plain undecorated xterm window. Thus you should pick one or more of the following lines:

services.xserver.desktopManager.kde4.enable = true;
services.xserver.desktopManager.xfce.enable = true;
services.xserver.windowManager.xmonad.enable = true;
services.xserver.windowManager.twm.enable = true;
services.xserver.windowManager.icewm.enable = true;

NixOS’s default display manager (the program that provides a graphical login prompt and manages the X server) is SLiM. You can select KDE’s kdm instead:

services.xserver.displayManager.kdm.enable = true;

The X server is started automatically at boot time. If you don’t want this to happen, you can set:

services.xserver.autorun = false;

The X server can then be started manually:

$ systemctl start display-manager.service

NVIDIA Graphics Cards

NVIDIA provides a proprietary driver for its graphics cards that has better 3D performance than the drivers. It is not enabled by default because it’s not free software. You can enable it as follows:

services.xserver.videoDrivers = [ "nvidia" ];

Or if you have an older card, you may have to use one of the legacy drivers:

services.xserver.videoDrivers = [ "nvidiaLegacy340" ];
services.xserver.videoDrivers = [ "nvidiaLegacy304" ];
services.xserver.videoDrivers = [ "nvidiaLegacy173" ];

You may need to reboot after enabling this driver to prevent a clash with other kernel modules.

On 64-bit systems, if you want full acceleration for 32-bit programs such as Wine, you should also set the following:

hardware.opengl.driSupport32Bit = true;

AMD Graphics Cards

AMD provides a proprietary driver for its graphics cards that has better 3D performance than the drivers. It is not enabled by default because it’s not free software. You can enable it as follows:

services.xserver.videoDrivers = [ "ati_unfree" ];

You will need to reboot after enabling this driver to prevent a clash with other kernel modules.

On 64-bit systems, if you want full acceleration for 32-bit programs such as Wine, you should also set the following:

hardware.opengl.driSupport32Bit = true;


Support for Synaptics touchpads (found in many laptops such as the Dell Latitude series) can be enabled as follows:

services.xserver.synaptics.enable = true;

The driver has many options (see Appendix A, Configuration Options). For instance, the following enables two-finger scrolling:

services.xserver.synaptics.twoFingerScroll = true;

Chapter 10. Networking

This section describes how to configure networking components on your NixOS machine.

10.1. NetworkManager

To facilitate network configuration, some desktop environments use NetworkManager. You can enable NetworkManager by setting:

networking.networkmanager.enable = true;

some desktop managers (e.g., GNOME) enable NetworkManager automatically for you.

All users that should have permission to change network settings must belong to the networkmanager group.

Note: networking.networkmanager and networking.wireless can not be enabled at the same time: you can still connect to the wireless networks using NetworkManager.

10.2. Secure Shell Access

Secure shell (SSH) access to your machine can be enabled by setting:

services.openssh.enable = true;

By default, root logins using a password are disallowed. They can be disabled entirely by setting services.openssh.permitRootLogin to "no".

You can declaratively specify authorised RSA/DSA public keys for a user as follows:

users.extraUsers.alice.openssh.authorizedKeys.keys =
  [ "ssh-dss AAAAB3NzaC1kc3MAAACBAPIkGWVEt4..." ];

10.3. IPv4 Configuration

By default, NixOS uses DHCP (specifically, dhcpcd) to automatically configure network interfaces. However, you can configure an interface manually as follows:

networking.interfaces.eth0.ip4 = [ { address = ""; prefixLength = 24; } ];

Typically you’ll also want to set a default gateway and set of name servers:

networking.defaultGateway = "";
networking.nameservers = [ "" ];

Note: Statically configured interfaces are set up by the systemd service interface-name-cfg.service. The default gateway and name server configuration is performed by network-setup.service.

The host name is set using networking.hostName:

networking.hostName = "cartman";

The default host name is nixos. Set it to the empty string ("") to allow the DHCP server to provide the host name.

10.4. IPv6 Configuration

IPv6 is enabled by default. Stateless address autoconfiguration is used to automatically assign IPv6 addresses to all interfaces. You can disable IPv6 support globally by setting:

networking.enableIPv6 = false;

10.5. Firewall

NixOS has a simple stateful firewall that blocks incoming connections and other unexpected packets. The firewall applies to both IPv4 and IPv6 traffic. It is enabled by default. It can be disabled as follows:

networking.firewall.enable = false;

If the firewall is enabled, you can open specific TCP ports to the outside world:

networking.firewall.allowedTCPPorts = [ 80 443 ];

Note that TCP port 22 (ssh) is opened automatically if the SSH daemon is enabled (services.openssh.enable = true). UDP ports can be opened through networking.firewall.allowedUDPPorts. Also of interest is

networking.firewall.allowPing = true;

to allow the machine to respond to ping requests. (ICMPv6 pings are always allowed.)

10.6. Wireless Networks

For a desktop installation using NetworkManager (e.g., GNOME), you just have to make sure the user is in the networkmanager group and you can skip the rest of this section on wireless networks.

NixOS will start wpa_supplicant for you if you enable this setting:

networking.wireless.enable = true;

NixOS currently does not generate wpa_supplicant's configuration file, /etc/wpa_supplicant.conf. You should edit this file yourself to define wireless networks, WPA keys and so on (see wpa_supplicant.conf(5)).

If you are using WPA2 the wpa_passphrase tool might be useful to generate the wpa_supplicant.conf.

$ wpa_passphrase ESSID PSK > /etc/wpa_supplicant.conf

After you have edited the wpa_supplicant.conf, you need to restart the wpa_supplicant service.

$ systemctl restart wpa_supplicant.service

10.7. Ad-Hoc Configuration

You can use networking.localCommands to specify shell commands to be run at the end of network-setup.service. This is useful for doing network configuration not covered by the existing NixOS modules. For instance, to statically configure an IPv6 address:

networking.localCommands =
    ip -6 addr add 2001:610:685:1::1/64 dev eth0

Chapter 11. Linux Kernel

You can override the Linux kernel and associated packages using the option boot.kernelPackages. For instance, this selects the Linux 3.10 kernel:

boot.kernelPackages = pkgs.linuxPackages_3_10;

Note that this not only replaces the kernel, but also packages that are specific to the kernel version, such as the NVIDIA video drivers. This ensures that driver packages are consistent with the kernel.

The default Linux kernel configuration should be fine for most users. You can see the configuration of your current kernel with the following command:

cat /proc/config.gz | gunzip

If you want to change the kernel configuration, you can use the packageOverrides feature (see Section 6.1.1, “Customising Packages”). For instance, to enable support for the kernel debugger KGDB:

nixpkgs.config.packageOverrides = pkgs:
  { linux_3_4 = pkgs.linux_3_4.override {
      extraConfig =
          KGDB y

extraConfig takes a list of Linux kernel configuration options, one per line. The name of the option should not include the prefix CONFIG_. The option value is typically y, n or m (to build something as a kernel module).

Kernel modules for hardware devices are generally loaded automatically by udev. You can force a module to be loaded via boot.kernelModules, e.g.

boot.kernelModules = [ "fuse" "kvm-intel" "coretemp" ];

If the module is required early during the boot (e.g. to mount the root file system), you can use boot.initrd.extraKernelModules:

boot.initrd.extraKernelModules = [ "cifs" ];

This causes the specified modules and their dependencies to be added to the initial ramdisk.

Kernel runtime parameters can be set through boot.kernel.sysctl, e.g.

boot.kernel.sysctl."net.ipv4.tcp_keepalive_time" = 120;

sets the kernel’s TCP keepalive time to 120 seconds. To see the available parameters, run sysctl -a.

Chapter 12. PostgreSQL

Source: modules/services/databases/postgresql.nix

Upstream documentation:

PostgreSQL is an advanced, free relational database.

12.1. Configuring

To enable PostgreSQL, add the following to your configuration.nix:

services.postgresql.enable = true;
services.postgresql.package = pkgs.postgresql94;

Note that you are required to specify the desired version of PostgreSQL (e.g. pkgs.postgresql94). Since upgrading your PostgreSQL version requires a database dump and reload (see below), NixOS cannot provide a default value for services.postgresql.package such as the most recent release of PostgreSQL.

By default, PostgreSQL stores its databases in /var/db/postgresql. You can override this using services.postgresql.dataDir, e.g.

services.postgresql.dataDir = "/data/postgresql";

12.2. Upgrading

FIXME: document dump/upgrade/load cycle.

12.3. Options

FIXME: auto-generated list of module options.

This chapter describes various aspects of managing a running NixOS system, such as how to use the systemd service manager.

Chapter 13. Service Management

In NixOS, all system services are started and monitored using the systemd program. Systemd is the “init” process of the system (i.e. PID 1), the parent of all other processes. It manages a set of so-called “units”, which can be things like system services (programs), but also mount points, swap files, devices, targets (groups of units) and more. Units can have complex dependencies; for instance, one unit can require that another unit must be successfully started before the first unit can be started. When the system boots, it starts a unit named; the dependencies of this unit cause all system services to be started, file systems to be mounted, swap files to be activated, and so on.

The command systemctl is the main way to interact with systemd. Without any arguments, it shows the status of active units:

$ systemctl
-.mount          loaded active mounted   /
swapfile.swap    loaded active active    /swapfile
sshd.service     loaded active running   SSH Daemon loaded active active    Graphical Interface

You can ask for detailed status information about a unit, for instance, the PostgreSQL database service:

$ systemctl status postgresql.service
postgresql.service - PostgreSQL Server
          Loaded: loaded (/nix/store/pn3q73mvh75gsrl8w7fdlfk3fq5qm5mw-unit/postgresql.service)
          Active: active (running) since Mon, 2013-01-07 15:55:57 CET; 9h ago
        Main PID: 2390 (postgres)
          CGroup: name=systemd:/system/postgresql.service
                  ├─2390 postgres
                  ├─2418 postgres: writer process
                  ├─2419 postgres: wal writer process
                  ├─2420 postgres: autovacuum launcher process
                  ├─2421 postgres: stats collector process
                  └─2498 postgres: zabbix zabbix [local] idle

Jan 07 15:55:55 hagbard postgres[2394]: [1-1] LOG:  database system was shut down at 2013-01-07 15:55:05 CET
Jan 07 15:55:57 hagbard postgres[2390]: [1-1] LOG:  database system is ready to accept connections
Jan 07 15:55:57 hagbard postgres[2420]: [1-1] LOG:  autovacuum launcher started
Jan 07 15:55:57 hagbard systemd[1]: Started PostgreSQL Server.

Note that this shows the status of the unit (active and running), all the processes belonging to the service, as well as the most recent log messages from the service.

Units can be stopped, started or restarted:

$ systemctl stop postgresql.service
$ systemctl start postgresql.service
$ systemctl restart postgresql.service

These operations are synchronous: they wait until the service has finished starting or stopping (or has failed). Starting a unit will cause the dependencies of that unit to be started as well (if necessary).

Chapter 14. Rebooting and Shutting Down

The system can be shut down (and automatically powered off) by doing:

$ shutdown

This is equivalent to running systemctl poweroff.

To reboot the system, run

$ reboot

which is equivalent to systemctl reboot. Alternatively, you can quickly reboot the system using kexec, which bypasses the BIOS by directly loading the new kernel into memory:

$ systemctl kexec

The machine can be suspended to RAM (if supported) using systemctl suspend, and suspended to disk using systemctl hibernate.

These commands can be run by any user who is logged in locally, i.e. on a virtual console or in X11; otherwise, the user is asked for authentication.

Chapter 15. User Sessions

Systemd keeps track of all users who are logged into the system (e.g. on a virtual console or remotely via SSH). The command loginctl allows querying and manipulating user sessions. For instance, to list all user sessions:

$ loginctl
   SESSION        UID USER             SEAT
        c1        500 eelco            seat0
        c3          0 root             seat0
        c4        500 alice

This shows that two users are logged in locally, while another is logged in remotely. (“Seats” are essentially the combinations of displays and input devices attached to the system; usually, there is only one seat.) To get information about a session:

$ loginctl session-status c3
c3 - root (0)
           Since: Tue, 2013-01-08 01:17:56 CET; 4min 42s ago
          Leader: 2536 (login)
            Seat: seat0; vc3
             TTY: /dev/tty3
         Service: login; type tty; class user
           State: online
          CGroup: name=systemd:/user/root/c3
                  ├─ 2536 /nix/store/10mn4xip9n7y9bxqwnsx7xwx2v2g34xn-shadow- --
                  ├─10339 -bash
                  └─10355 w3m

This shows that the user is logged in on virtual console 3. It also lists the processes belonging to this session. Since systemd keeps track of this, you can terminate a session in a way that ensures that all the session’s processes are gone:

$ loginctl terminate-session c3

Chapter 16. Control Groups

To keep track of the processes in a running system, systemd uses control groups (cgroups). A control group is a set of processes used to allocate resources such as CPU, memory or I/O bandwidth. There can be multiple control group hierarchies, allowing each kind of resource to be managed independently.

The command systemd-cgls lists all control groups in the systemd hierarchy, which is what systemd uses to keep track of the processes belonging to each service or user session:

$ systemd-cgls
│ └─eelco
│   └─c1
│     ├─ 2567 -:0
│     ├─ 2682 kdeinit4: kdeinit4 Running...
│     ├─ ...
│     └─10851 sh -c less -R
  │ ├─2444 httpd -f /nix/store/3pyacby5cpr55a03qwbnndizpciwq161-httpd.conf -DNO_DETACH
  │ └─...
  │ └─2376 dhcpcd --config /nix/store/f8dif8dsi2yaa70n03xir8r653776ka6-dhcpcd.conf
  └─ ...

Similarly, systemd-cgls cpu shows the cgroups in the CPU hierarchy, which allows per-cgroup CPU scheduling priorities. By default, every systemd service gets its own CPU cgroup, while all user sessions are in the top-level CPU cgroup. This ensures, for instance, that a thousand run-away processes in the httpd.service cgroup cannot starve the CPU for one process in the postgresql.service cgroup. (By contrast, it they were in the same cgroup, then the PostgreSQL process would get 1/1001 of the cgroup’s CPU time.) You can limit a service’s CPU share in configuration.nix: = 512;

By default, every cgroup has 1024 CPU shares, so this will halve the CPU allocation of the httpd.service cgroup.

There also is a memory hierarchy that controls memory allocation limits; by default, all processes are in the top-level cgroup, so any service or session can exhaust all available memory. Per-cgroup memory limits can be specified in configuration.nix; for instance, to limit httpd.service to 512 MiB of RAM (excluding swap): = "512M";

The command systemd-cgtop shows a continuously updated list of all cgroups with their CPU and memory usage.

Chapter 17. Logging

System-wide logging is provided by systemd’s journal, which subsumes traditional logging daemons such as syslogd and klogd. Log entries are kept in binary files in /var/log/journal/. The command journalctl allows you to see the contents of the journal. For example,

$ journalctl -b

shows all journal entries since the last reboot. (The output of journalctl is piped into less by default.) You can use various options and match operators to restrict output to messages of interest. For instance, to get all messages from PostgreSQL:

$ journalctl -u postgresql.service
-- Logs begin at Mon, 2013-01-07 13:28:01 CET, end at Tue, 2013-01-08 01:09:57 CET. --
Jan 07 15:44:14 hagbard postgres[2681]: [2-1] LOG:  database system is shut down
-- Reboot --
Jan 07 15:45:10 hagbard postgres[2532]: [1-1] LOG:  database system was shut down at 2013-01-07 15:44:14 CET
Jan 07 15:45:13 hagbard postgres[2500]: [1-1] LOG:  database system is ready to accept connections

Or to get all messages since the last reboot that have at least a “critical” severity level:

$ journalctl -b -p crit
Dec 17 21:08:06 mandark sudo[3673]: pam_unix(sudo:auth): auth could not identify password for [alice]
Dec 29 01:30:22 mandark kernel[6131]: [1053513.909444] CPU6: Core temperature above threshold, cpu clock throttled (total events = 1)

The system journal is readable by root and by users in the wheel and systemd-journal groups. All users have a private journal that can be read using journalctl.

Chapter 18. Cleaning the Nix Store

Nix has a purely functional model, meaning that packages are never upgraded in place. Instead new versions of packages end up in a different location in the Nix store (/nix/store). You should periodically run Nix’s garbage collector to remove old, unreferenced packages. This is easy:

$ nix-collect-garbage

Alternatively, you can use a systemd unit that does the same in the background:

$ systemctl start nix-gc.service

You can tell NixOS in configuration.nix to run this unit automatically at certain points in time, for instance, every night at 03:15:

nix.gc.automatic = true;
nix.gc.dates = "03:15";

The commands above do not remove garbage collector roots, such as old system configurations. Thus they do not remove the ability to roll back to previous configurations. The following command deletes old roots, removing the ability to roll back to them:

$ nix-collect-garbage -d

You can also do this for specific profiles, e.g.

$ nix-env -p /nix/var/nix/profiles/per-user/eelco/profile --delete-generations old

Note that NixOS system configurations are stored in the profile /nix/var/nix/profiles/system.

Another way to reclaim disk space (often as much as 40% of the size of the Nix store) is to run Nix’s store optimiser, which seeks out identical files in the store and replaces them with hard links to a single copy.

$ nix-store --optimise

Since this command needs to read the entire Nix store, it can take quite a while to finish.

Chapter 19. Container Management

NixOS allows you to easily run other NixOS instances as containers. Containers are a light-weight approach to virtualisation that runs software in the container at the same speed as in the host system. NixOS containers share the Nix store of the host, making container creation very efficient.

Warning: Currently, NixOS containers are not perfectly isolated from the host system. This means that a user with root access to the container can do things that affect the host. So you should not give container root access to untrusted users.

NixOS containers can be created in two ways: imperatively, using the command nixos-container, and declaratively, by specifying them in your configuration.nix. The declarative approach implies that containers get upgraded along with your host system when you run nixos-rebuild, which is often not what you want. By contrast, in the imperative approach, containers are configured and updated independently from the host system.

19.1. Imperative Container Management

We’ll cover imperative container management using nixos-container first. You create a container with identifier foo as follows:

$ nixos-container create foo

This creates the container’s root directory in /var/lib/containers/foo and a small configuration file in /etc/containers/foo.conf. It also builds the container’s initial system configuration and stores it in /nix/var/nix/profiles/per-container/foo/system. You can modify the initial configuration of the container on the command line. For instance, to create a container that has sshd running, with the given public key for root:

$ nixos-container create foo --config 'services.openssh.enable = true; \
  users.extraUsers.root.openssh.authorizedKeys.keys = ["ssh-dss AAAAB3N…"];'

Creating a container does not start it. To start the container, run:

$ nixos-container start foo

This command will return as soon as the container has booted and has reached On the host, the container runs within a systemd unit called container@container-name.service. Thus, if something went wrong, you can get status info using systemctl:

$ systemctl status container@foo

If the container has started succesfully, you can log in as root using the root-login operation:

$ nixos-container root-login foo

Note that only root on the host can do this (since there is no authentication). You can also get a regular login prompt using the login operation, which is available to all users on the host:

$ nixos-container login foo
foo login: alice
Password: ***

With nixos-container run, you can execute arbitrary commands in the container:

$ nixos-container run foo -- uname -a
Linux foo 3.4.82 #1-NixOS SMP Thu Mar 20 14:44:05 UTC 2014 x86_64 GNU/Linux

There are several ways to change the configuration of the container. First, on the host, you can edit /var/lib/container/name/etc/nixos/configuration.nix, and run

$ nixos-container update foo

This will build and activate the new configuration. You can also specify a new configuration on the command line:

$ nixos-container update foo --config 'services.httpd.enable = true; \
  services.httpd.adminAddr = "";'

$ curl http://$(nixos-container show-ip foo)/

However, note that this will overwrite the container’s /etc/nixos/configuration.nix.

Alternatively, you can change the configuration from within the container itself by running nixos-rebuild switch inside the container. Note that the container by default does not have a copy of the NixOS channel, so you should run nix-channel --update first.

Containers can be stopped and started using nixos-container stop and nixos-container start, respectively, or by using systemctl on the container’s service unit. To destroy a container, including its file system, do

$ nixos-container destroy foo

19.2. Declarative Container Specification

You can also specify containers and their configuration in the host’s configuration.nix. For example, the following specifies that there shall be a container named database running PostgreSQL:

containers.database =
  { config =
      { config, pkgs, ... }:
      { services.postgresql.enable = true;
        services.postgresql.package = pkgs.postgresql92;

If you run nixos-rebuild switch, the container will be built and started. If the container was already running, it will be updated in place, without rebooting.

By default, declarative containers share the network namespace of the host, meaning that they can listen on (privileged) ports. However, they cannot change the network configuration. You can give a container its own network as follows:

containers.database =
  { privateNetwork = true;
    hostAddress = "";
    localAddress = "";

This gives the container a private virtual Ethernet interface with IP address, which is hooked up to a virtual Ethernet interface on the host with IP address (See the next section for details on container networking.)

To disable the container, just remove it from configuration.nix and run nixos-rebuild switch. Note that this will not delete the root directory of the container in /var/lib/containers.

Declarative containers can be started and stopped using the corresponding systemd service, e.g. systemctl start container@database.

19.3. Container Networking

When you create a container using nixos-container create, it gets it own private IPv4 address in the range You can get the container’s IPv4 address as follows:

$ nixos-container show-ip foo

$ ping -c1
64 bytes from icmp_seq=1 ttl=64 time=0.106 ms

Networking is implemented using a pair of virtual Ethernet devices. The network interface in the container is called eth0, while the matching interface in the host is called ve-container-name (e.g., ve-foo). The container has its own network namespace and the CAP_NET_ADMIN capability, so it can perform arbitrary network configuration such as setting up firewall rules, without affecting or having access to the host’s network.

By default, containers cannot talk to the outside network. If you want that, you should set up Network Address Translation (NAT) rules on the host to rewrite container traffic to use your external IP address. This can be accomplished using the following configuration on the host:

networking.nat.enable = true;
networking.nat.internalInterfaces = ["ve-+"];
networking.nat.externalInterface = "eth0";

where eth0 should be replaced with the desired external interface. Note that ve-+ is a wildcard that matches all container interfaces.

Chapter 20. Troubleshooting

This chapter describes solutions to common problems you might encounter when you manage your NixOS system.

20.1. Boot Problems

If NixOS fails to boot, there are a number of kernel command line parameters that may help you to identify or fix the issue. You can add these parameters in the GRUB boot menu by pressing “e” to modify the selected boot entry and editing the line starting with linux. The following are some useful kernel command line parameters that are recognised by the NixOS boot scripts or by systemd:


Start a root shell if something goes wrong in stage 1 of the boot process (the initial ramdisk). This is disabled by default because there is no authentication for the root shell.


Start an interactive shell in stage 1 before anything useful has been done. That is, no modules have been loaded and no file systems have been mounted, except for /proc and /sys.


Print every shell command executed by the stage 1 and 2 boot scripts.


Boot into rescue mode (a.k.a. single user mode). This will cause systemd to start nothing but the unit, which runs sulogin to prompt for the root password and start a root login shell. Exiting the shell causes the system to continue with the normal boot process.

systemd.log_level=debug systemd.log_target=console

Make systemd very verbose and send log messages to the console instead of the journal.

For more parameters recognised by systemd, see systemd(1).

If no login prompts or X11 login screens appear (e.g. due to hanging dependencies), you can press Alt+ArrowUp. If you’re lucky, this will start rescue mode (described above). (Also note that since most units have a 90-second timeout before systemd gives up on them, the agetty login prompts should appear eventually unless something is very wrong.)

20.2. Maintenance Mode

You can enter rescue mode by running:

$ systemctl rescue

This will eventually give you a single-user root shell. Systemd will stop (almost) all system services. To get out of maintenance mode, just exit from the rescue shell.

20.3. Rolling Back Configuration Changes

After running nixos-rebuild to switch to a new configuration, you may find that the new configuration doesn’t work very well. In that case, there are several ways to return to a previous configuration.

First, the GRUB boot manager allows you to boot into any previous configuration that hasn’t been garbage-collected. These configurations can be found under the GRUB submenu “NixOS - All configurations”. This is especially useful if the new configuration fails to boot. After the system has booted, you can make the selected configuration the default for subsequent boots:

$ /run/current-system/bin/switch-to-configuration boot

Second, you can switch to the previous configuration in a running system:

$ nixos-rebuild switch --rollback

This is equivalent to running:

$ /nix/var/nix/profiles/system-N-link/bin/switch-to-configuration switch

where N is the number of the NixOS system configuration. To get a list of the available configurations, do:

$ ls -l /nix/var/nix/profiles/system-*-link
lrwxrwxrwx 1 root root 78 Aug 12 13:54 /nix/var/nix/profiles/system-268-link -> /nix/store/202b...-nixos-13.07pre4932_5a676e4-4be1055

20.4. Nix Store Corruption

After a system crash, it’s possible for files in the Nix store to become corrupted. (For instance, the Ext4 file system has the tendency to replace un-synced files with zero bytes.) NixOS tries hard to prevent this from happening: it performs a sync before switching to a new configuration, and Nix’s database is fully transactional. If corruption still occurs, you may be able to fix it automatically.

If the corruption is in a path in the closure of the NixOS system configuration, you can fix it by doing

$ nixos-rebuild switch --repair

This will cause Nix to check every path in the closure, and if its cryptographic hash differs from the hash recorded in Nix’s database, the path is rebuilt or redownloaded.

You can also scan the entire Nix store for corrupt paths:

$ nix-store --verify --check-contents --repair

Any corrupt paths will be redownloaded if they’re available in a binary cache; otherwise, they cannot be repaired.

20.5. Network Problems

Nix uses a so-called binary cache to optimise building a package from source into downloading it as a pre-built binary. That is, whenever a command like nixos-rebuild needs a path in the Nix store, Nix will try to download that path from the Internet rather than build it from source. The default binary cache is If this cache is unreachable, Nix operations may take a long time due to HTTP connection timeouts. You can disable the use of the binary cache by adding --option use-binary-caches false, e.g.

$ nixos-rebuild switch --option use-binary-caches false

If you have an alternative binary cache at your disposal, you can use it instead:

$ nixos-rebuild switch --option binary-caches

This chapter describes how you can modify and extend NixOS.

Chapter 21. Getting the Sources

By default, NixOS’s nixos-rebuild command uses the NixOS and Nixpkgs sources provided by the nixos-unstable channel (kept in /nix/var/nix/profiles/per-user/root/channels/nixos). To modify NixOS, however, you should check out the latest sources from Git. This is done using the following command:

$ nixos-checkout /my/sources


$ mkdir -p /my/sources
$ cd /my/sources
$ nix-env -i git
$ git clone git://
$ cd nixpkgs
$ git remote add channels git://
$ git remote update channels

This will check out the latest NixOS sources to /my/sources/nixpkgs/nixos and the Nixpkgs sources to /my/sources/nixpkgs. (The NixOS source tree lives in a subdirectory of the Nixpkgs repository.) The remote channels refers to a read-only repository that tracks the Nixpkgs/NixOS channels (see Chapter 4, Upgrading NixOS for more information about channels). Thus, the Git branch channels/nixos-14.12 will contain the latest built and tested version available in the nixos-14.12 channel.

It’s often inconvenient to develop directly on the master branch, since if somebody has just committed (say) a change to GCC, then the binary cache may not have caught up yet and you’ll have to rebuild everything from source. So you may want to create a local branch based on your current NixOS version:

$ nixos-version
14.04.273.ea1952b (Baboon)

$ git checkout -b local ea1952b

Or, to base your local branch on the latest version available in a NixOS channel:

$ git remote update channels
$ git checkout -b local channels/nixos-14.12

(Replace nixos-14.12 with the name of the channel you want to use.) You can use git merge or git rebase to keep your local branch in sync with the channel, e.g.

$ git remote update channels
$ git merge channels/nixos-14.12

You can use git cherry-pick to copy commits from your local branch to the upstream branch.

If you want to rebuild your system using your (modified) sources, you need to tell nixos-rebuild about them using the -I flag:

$ nixos-rebuild switch -I nixpkgs=/my/sources/nixpkgs

If you want nix-env to use the expressions in /my/sources, use nix-env -f /my/sources/nixpkgs, or change the default by adding a symlink in ~/.nix-defexpr:

$ ln -s /my/sources/nixpkgs ~/.nix-defexpr/nixpkgs

You may want to delete the symlink ~/.nix-defexpr/channels_root to prevent root’s NixOS channel from clashing with your own tree.

Chapter 22. Writing NixOS Modules

NixOS has a modular system for declarative configuration. This system combines multiple modules to produce the full system configuration. One of the modules that constitute the configuration is /etc/nixos/configuration.nix. Most of the others live in the nixos/modules subdirectory of the Nixpkgs tree.

Each NixOS module is a file that handles one logical aspect of the configuration, such as a specific kind of hardware, a service, or network settings. A module configuration does not have to handle everything from scratch; it can use the functionality provided by other modules for its implementation. Thus a module can declare options that can be used by other modules, and conversely can define options provided by other modules in its own implementation. For example, the module pam.nix declares the option that allows other modules (e.g. sshd.nix) to define PAM services; and it defines the option environment.etc (declared by etc.nix) to cause files to be created in /etc/pam.d.

In Chapter 5, Configuration Syntax, we saw the following structure of NixOS modules:

{ config, pkgs, ... }:

{ option definitions

This is actually an abbreviated form of module that only defines options, but does not declare any. The structure of full NixOS modules is shown in Example 22.1, “Structure of NixOS Modules”.

Example 22.1. Structure of NixOS Modules

{ config, pkgs, ... }: 1

  imports =
    [ paths of other modules 2

  options = {
    option declarations 3

  config = {
    option definitions 4

The meaning of each part is as follows.


This line makes the current Nix expression a function. The variable pkgs contains Nixpkgs, while config contains the full system configuration. This line can be omitted if there is no reference to pkgs and config inside the module.


This list enumerates the paths to other NixOS modules that should be included in the evaluation of the system configuration. A default set of modules is defined in the file modules/module-list.nix. These don't need to be added in the import list.


The attribute options is a nested set of option declarations (described below).


The attribute config is a nested set of option definitions (also described below).

Example 22.2, “NixOS Module for the “locate” Service” shows a module that handles the regular update of the “locate” database, an index of all files in the file system. This module declares two options that can be defined by other modules (typically the user’s configuration.nix): services.locate.enable (whether the database should be updated) and services.locate.period (when the update should be done). It implements its functionality by defining two options declared by other modules: (the set of all systemd services) and services.cron.systemCronJobs (the list of commands to be executed periodically by cron).

Example 22.2. NixOS Module for the “locate” Service

{ config, lib, pkgs, ... }:

with lib;

let locatedb = "/var/cache/locatedb"; in

  options = {

    services.locate = {

      enable = mkOption {
        type = types.bool;
        default = false;
        description = ''
          If enabled, NixOS will periodically update the database of
          files used by the locate command.

      period = mkOption {
        type = types.str;
        default = "15 02 * * *";
        description = ''
          This option defines (in the format used by cron) when the
          locate database is updated.  The default is to update at
          02:15 at night every day.



  config = { =
      { description = "Update Locate Database";
        path  = [ ];
        script =
            mkdir -m 0755 -p $(dirname ${locatedb})
            exec updatedb --localuser=nobody --output=${locatedb} --prunepaths='/tmp /var/tmp /run'

    services.cron.systemCronJobs = optional
      "${} root ${config.systemd.package}/bin/systemctl start update-locatedb.service";


22.1. Option Declarations

An option declaration specifies the name, type and description of a NixOS configuration option. It is illegal to define an option that hasn’t been declared in any module. A option declaration generally looks like this:

options = {
  name = mkOption {
    type = type specification;
    default = default value;
    example = example value;
    description = "Description for use in the NixOS manual.";

The function mkOption accepts the following arguments.


The type of the option (see below). It may be omitted, but that’s not advisable since it may lead to errors that are hard to diagnose.


The default value used if no value is defined by any module. A default is not required; in that case, if the option value is ever used, an error will be thrown.


An example value that will be shown in the NixOS manual.


A textual description of the option, in DocBook format, that will be included in the NixOS manual.

Here is a non-exhaustive list of option types:


A Boolean.

An integer.


A string.


A string. If there are multiple definitions, they are concatenated, with newline characters in between.


A path, defined as anything that, when coerced to a string, starts with a slash. This includes derivations.


A derivation (such as pkgs.hello) or a store path (such as /nix/store/1ifi1cfbfs5iajmvwgrbmrnrw3a147h9-hello-2.10).

types.listOf t

A list of elements of type t (e.g., types.listOf types.str is a list of strings). Multiple definitions are concatenated together.

types.attrsOf t

A set of elements of type t (e.g., types.attrsOf is a set of name/value pairs, the values being integers).

types.nullOr t

Either the value null or something of type t.

You can also create new types using the function mkOptionType. See lib/types.nix in Nixpkgs for details.

22.2. Option Definitions

Option definitions are generally straight-forward bindings of values to option names, like

config = {
  services.httpd.enable = true;

However, sometimes you need to wrap an option definition or set of option definitions in a property to achieve certain effects:

Delaying Conditionals

If a set of option definitions is conditional on the value of another option, you may need to use mkIf. Consider, for instance:

config = if then {
  environment.systemPackages = [ ... ];
} else {};

This definition will cause Nix to fail with an “infinite recursion” error. Why? Because the value of depends on the value being constructed here. After all, you could also write the clearly circular and contradictory:

config = if then {
  services.httpd.enable = false;
} else {
  services.httpd.enable = true;

The solution is to write:

config = mkIf {
  environment.systemPackages = [ ... ];

The special function mkIf causes the evaluation of the conditional to be “pushed down” into the individual definitions, as if you had written:

config = {
  environment.systemPackages = if then [ ... ] else [];

Setting Priorities

A module can override the definitions of an option in other modules by setting a priority. All option definitions that do not have the lowest priority value are discarded. By default, option definitions have priority 1000. You can specify an explicit priority by using mkOverride, e.g.

services.openssh.enable = mkOverride 10 false;

This definition causes all other definitions with priorities above 10 to be discarded. The function mkForce is equal to mkOverride 50.

Merging Configurations

In conjunction with mkIf, it is sometimes useful for a module to return multiple sets of option definitions, to be merged together as if they were declared in separate modules. This can be done using mkMerge:

config = mkMerge
  [ # Unconditional stuff.
    { environment.systemPackages = [ ... ];
    # Conditional stuff.
    (mkIf {
      environment.systemPackages = [ ... ];

Chapter 23. Building Specific Parts of NixOS

With the command nix-build, you can build specific parts of your NixOS configuration. This is done as follows:

$ cd /path/to/nixpkgs/nixos
$ nix-build -A config.option

where option is a NixOS option with type “derivation” (i.e. something that can be built). Attributes of interest include:

The top-level option that builds the entire NixOS system. Everything else in your configuration is indirectly pulled in by this option. This is what nixos-rebuild builds and what /run/current-system points to afterwards.

A shortcut to build this is:

$ nix-build -A system

The NixOS manual.

A tree of symlinks that form the static parts of /etc.,

The initial ramdisk and kernel of the system. This allows a quick way to test whether the kernel and the initial ramdisk boot correctly, by using QEMU’s -kernel and -initrd options:

$ nix-build -A -o initrd
$ nix-build -A -o kernel
$ qemu-system-x86_64 -kernel ./kernel/bzImage -initrd ./initrd/initrd -hda /dev/null,,

These build the corresponding NixOS commands.


This builds the unit with the specified name. Note that since unit names contain dots (e.g. httpd.service), you need to put them between quotes, like this:

$ nix-build -A 'config.systemd.units."httpd.service".unit'

You can also test individual units, without rebuilding the whole system, by putting them in /run/systemd/system:

$ cp $(nix-build -A 'config.systemd.units."httpd.service".unit')/httpd.service \
$ systemctl daemon-reload
$ systemctl start tmp-httpd.service

Note that the unit must not have the same name as any unit in /etc/systemd/system since those take precedence over /run/systemd/system. That’s why the unit is installed as tmp-httpd.service here.

Chapter 24. Building Your Own NixOS CD

Building a NixOS CD is as easy as configuring your own computer. The idea is to use another module which will replace your configuration.nix to configure the system that would be installed on the CD.

Default CD/DVD configurations are available inside nixos/modules/installer/cd-dvd. To build them you have to set NIXOS_CONFIG before running nix-build to build the ISO.

$ nix-build -A -I nixos-config=modules/installer/cd-dvd/installation-cd-minimal.nix

Before burning your CD/DVD, you can check the content of the image by mounting anywhere like suggested by the following command:

$ mount -o loop -t iso9660 ./result/iso/cd.iso /mnt/iso

Chapter 25. NixOS Tests

When you add some feature to NixOS, you should write a test for it. NixOS tests are kept in the directory nixos/tests, and are executed (using Nix) by a testing framework that automatically starts one or more virtual machines containing the NixOS system(s) required for the test.

25.1. Writing Tests

A NixOS test is a Nix expression that has the following structure:

import ./make-test.nix {

  # Either the configuration of a single machine:
  machine =
    { config, pkgs, ... }:
    { configuration…

  # Or a set of machines:
  nodes =
    { machine1 =
        { config, pkgs, ... }: {  };
      machine2 =
        { config, pkgs, ... }: {  };

  testScript =
      Perl code…

The attribute testScript is a bit of Perl code that executes the test (described below). During the test, it will start one or more virtual machines, the configuration of which is described by the attribute machine (if you need only one machine in your test) or by the attribute nodes (if you need multiple machines). For instance, login.nix only needs a single machine to test whether users can log in on the virtual console, whether device ownership is correctly maintained when switching between consoles, and so on. On the other hand, nfs.nix, which tests NFS client and server functionality in the Linux kernel (including whether locks are maintained across server crashes), requires three machines: a server and two clients.

There are a few special NixOS configuration options for test VMs:


The memory of the VM in megabytes.


The virtual networks to which the VM is connected. See nat.nix for an example.


By default, the Nix store in the VM is not writable. If you enable this option, a writable union file system is mounted on top of the Nix store to make it appear writable. This is necessary for tests that run Nix operations that modify the store.

For more options, see the module qemu-vm.nix.

The test script is a sequence of Perl statements that perform various actions, such as starting VMs, executing commands in the VMs, and so on. Each virtual machine is represented as an object stored in the variable $name, where name is the identifier of the machine (which is just machine if you didn’t specify multiple machines using the nodes attribute). For instance, the following starts the machine, waits until it has finished booting, then executes a command and checks that the output is more-or-less correct:

$machine->succeed("uname") =~ /Linux/;

The first line is actually unnecessary; machines are implicitly started when you first execute an action on them (such as waitForUnit or succeed). If you have multiple machines, you can speed up the test by starting them in parallel:


The following methods are available on machine objects:


Start the virtual machine. This method is asynchronous — it does not wait for the machine to finish booting.


Shut down the machine, waiting for the VM to exit.


Simulate a sudden power failure, by telling the VM to exit immediately.


Simulate unplugging the Ethernet cable that connects the machine to the other machines.


Undo the effect of block.


Take a picture of the display of the virtual machine, in PNG format. The screenshot is linked from the HTML log.


Return a textual representation of what is currently visible on the machine's screen using optical character recognition.

Note: This requires passing enableOCR to the test attribute set.

Send a command to the QEMU monitor. This is rarely used, but allows doing stuff such as attaching virtual USB disks to a running machine.


Simulate pressing keys on the virtual keyboard, e.g., sendKeys("ctrl-alt-delete").


Simulate typing a sequence of characters on the virtual keyboard, e.g., sendKeys("foobar\n") will type the string foobar followed by the Enter key.


Execute a shell command, returning a list (status, stdout).


Execute a shell command, raising an exception if the exit status is not zero, otherwise returning the standard output.


Like succeed, but raising an exception if the command returns a zero status.


Repeat a shell command with 1-second intervals until it succeeds.


Repeat a shell command with 1-second intervals until it fails.


Wait until the specified systemd unit has reached the “active” state.


Wait until the specified file exists.


Wait until a process is listening on the given TCP port (on localhost, at least).


Wait until nobody is listening on the given TCP port.


Wait until the X11 server is accepting connections.


Wait until the supplied regular expressions matches the textual contents of the screen by using optical character recognition (see getScreenText).

Note: This requires passing enableOCR to the test attribute set.

Wait until an X11 window has appeared whose name matches the given regular expression, e.g., waitForWindow(qr/Terminal/).

25.2. Running Tests

You can run tests using nix-build. For example, to run the test login.nix, you just do:

$ nix-build '<nixpkgs/nixos/tests/login.nix>'

or, if you don’t want to rely on NIX_PATH:

$ cd /my/nixpkgs/nixos/tests
$ nix-build login.nix
running the VM test script
machine: QEMU running (pid 8841)
6 out of 6 tests succeeded

After building/downloading all required dependencies, this will perform a build that starts a QEMU/KVM virtual machine containing a NixOS system. The virtual machine mounts the Nix store of the host; this makes VM creation very fast, as no disk image needs to be created. Afterwards, you can view a pretty-printed log of the test:

$ firefox result/log.html

25.3. Running Tests interactively

The test itself can be run interactively. This is particularly useful when developing or debugging a test:

$ nix-build nixos/tests/login.nix -A driver
$ ./result/bin/nixos-test-driver
starting VDE switch for network 1

You can then take any Perl statement, e.g.

> startAll
> testScript
> $machine->succeed("touch /tmp/foo")

The function testScript executes the entire test script and drops you back into the test driver command line upon its completion. This allows you to inspect the state of the VMs after the test (e.g. to debug the test script).

To just start and experiment with the VMs, run:

$ nix-build nixos/tests/login.nix -A driver
$ ./result/bin/nixos-run-vms

The script nixos-run-vms starts the virtual machines defined by test. The root file system of the VMs is created on the fly and kept across VM restarts in ./hostname.qcow2.

Chapter 26. Testing the Installer

Building, burning, and booting from an installation CD is rather tedious, so here is a quick way to see if the installer works properly:

$ nix-build -A
$ mount -t tmpfs none /mnt
$ ./result/bin/nixos-install

To start a login shell in the new NixOS installation in /mnt:

$ ./result/bin/nixos-install --chroot