lix/doc/manual/command-ref/conf-file.xml

<refentry xmlns="http://docbook.org/ns/docbook"
          xmlns:xlink="http://www.w3.org/1999/xlink"
          xmlns:xi="http://www.w3.org/2001/XInclude"
          xml:id="sec-conf-file">

<refmeta>
  <refentrytitle>nix.conf</refentrytitle>
  <manvolnum>5</manvolnum>
  <refmiscinfo class="source">Nix</refmiscinfo>
  <refmiscinfo class="version"><xi:include href="../version.txt" parse="text"/></refmiscinfo>
</refmeta>

<refnamediv>
  <refname>nix.conf</refname>
  <refpurpose>Nix configuration file</refpurpose>
</refnamediv>

<refsection><title>Description</title>

<para>Nix reads settings from two configuration files:</para>

<itemizedlist>

  <listitem>
    <para>The system-wide configuration file
    <filename><replaceable>sysconfdir</replaceable>/nix/nix.conf</filename>
    (i.e. <filename>/etc/nix/nix.conf</filename> on most systems), or
    <filename>$NIX_CONF_DIR/nix.conf</filename> if
    <envar>NIX_CONF_DIR</envar> is set.</para>
  </listitem>

  <listitem>
    <para>The user configuration file
    <filename>$XDG_CONFIG_HOME/nix/nix.conf</filename>, or
    <filename>~/.config/nix/nix.conf</filename> if
    <envar>XDG_CONFIG_HOME</envar> is not set.</para>
  </listitem>

</itemizedlist>

<para>The configuration files consist of
<literal><replaceable>name</replaceable> =
<replaceable>value</replaceable></literal> pairs, one per line.
Comments start with a <literal>#</literal> character.  Here is an
example configuration file:</para>

<programlisting>
keep-outputs = true       # Nice for developers
keep-derivations = true   # Idem
</programlisting>

<para>You can override settings on the command line using the
<option>--option</option> flag, e.g. <literal>--option keep-outputs
false</literal>.</para>

<para>The following settings are currently available:

<variablelist>


  <varlistentry xml:id="conf-keep-outputs"><term><literal>keep-outputs</literal></term>

    <listitem><para>If <literal>true</literal>, the garbage collector
    will keep the outputs of non-garbage derivations.  If
    <literal>false</literal> (default), outputs will be deleted unless
    they are GC roots themselves (or reachable from other roots).</para>

    <para>In general, outputs must be registered as roots separately.
    However, even if the output of a derivation is registered as a
    root, the collector will still delete store paths that are used
    only at build time (e.g., the C compiler, or source tarballs
    downloaded from the network).  To prevent it from doing so, set
    this option to <literal>true</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-keep-derivations"><term><literal>keep-derivations</literal></term>

    <listitem><para>If <literal>true</literal> (default), the garbage
    collector will keep the derivations from which non-garbage store
    paths were built.  If <literal>false</literal>, they will be
    deleted unless explicitly registered as a root (or reachable from
    other roots).</para>

    <para>Keeping derivation around is useful for querying and
    traceability (e.g., it allows you to ask with what dependencies or
    options a store path was built), so by default this option is on.
    Turn it off to save a bit of disk space (or a lot if
    <literal>keep-outputs</literal> is also turned on).</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>keep-env-derivations</literal></term>

    <listitem><para>If <literal>false</literal> (default), derivations
    are not stored in Nix user environments.  That is, the derivation
    any build-time-only dependencies may be garbage-collected.</para>

    <para>If <literal>true</literal>, when you add a Nix derivation to
    a user environment, the path of the derivation is stored in the
    user environment.  Thus, the derivation will not be
    garbage-collected until the user environment generation is deleted
    (<command>nix-env --delete-generations</command>).  To prevent
    build-time-only dependencies from being collected, you should also
    turn on <literal>keep-outputs</literal>.</para>

    <para>The difference between this option and
    <literal>keep-derivations</literal> is that this one is
    “sticky”: it applies to any user environment created while this
    option was enabled, while <literal>keep-derivations</literal>
    only applies at the moment the garbage collector is
    run.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-max-jobs"><term><literal>max-jobs</literal></term>

    <listitem><para>This option defines the maximum number of jobs
    that Nix will try to build in parallel.  The default is
    <literal>1</literal>. The special value <literal>auto</literal>
    causes Nix to use the number of CPUs in your system.  It can be
    overridden using the <option
    linkend='opt-max-jobs'>--max-jobs</option> (<option>-j</option>)
    command line switch.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-cores"><term><literal>cores</literal></term>

    <listitem><para>Sets the value of the
    <envar>NIX_BUILD_CORES</envar> environment variable in the
    invocation of builders.  Builders can use this variable at their
    discretion to control the maximum amount of parallelism.  For
    instance, in Nixpkgs, if the derivation attribute
    <varname>enableParallelBuilding</varname> is set to
    <literal>true</literal>, the builder passes the
    <option>-j<replaceable>N</replaceable></option> flag to GNU Make.
    It can be overridden using the <option
    linkend='opt-cores'>--cores</option> command line switch and
    defaults to <literal>1</literal>.  The value <literal>0</literal>
    means that the builder should use all available CPU cores in the
    system.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-max-silent-time"><term><literal>max-silent-time</literal></term>

    <listitem>

      <para>This option defines the maximum number of seconds that a
      builder can go without producing any data on standard output or
      standard error.  This is useful (for instance in an automated
      build system) to catch builds that are stuck in an infinite
      loop, or to catch remote builds that are hanging due to network
      problems.  It can be overridden using the <option
      linkend="opt-max-silent-time">--max-silent-time</option> command
      line switch.</para>

      <para>The value <literal>0</literal> means that there is no
      timeout.  This is also the default.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-timeout"><term><literal>timeout</literal></term>

    <listitem>

      <para>This option defines the maximum number of seconds that a
      builder can run.  This is useful (for instance in an automated
      build system) to catch builds that are stuck in an infinite loop
      but keep writing to their standard output or standard error.  It
      can be overridden using the <option
      linkend="opt-timeout">--timeout</option> command line
      switch.</para>

      <para>The value <literal>0</literal> means that there is no
      timeout.  This is also the default.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-max-build-log-size"><term><literal>max-build-log-size</literal></term>

    <listitem>

      <para>This option defines the maximum number of bytes that a
      builder can write to its stdout/stderr.  If the builder exceeds
      this limit, it’s killed.  A value of <literal>0</literal> (the
      default) means that there is no limit.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-build-users-group"><term><literal>build-users-group</literal></term>

    <listitem><para>This options specifies the Unix group containing
    the Nix build user accounts.  In multi-user Nix installations,
    builds should not be performed by the Nix account since that would
    allow users to arbitrarily modify the Nix store and database by
    supplying specially crafted builders; and they cannot be performed
    by the calling user since that would allow him/her to influence
    the build result.</para>

    <para>Therefore, if this option is non-empty and specifies a valid
    group, builds will be performed under the user accounts that are a
    member of the group specified here (as listed in
    <filename>/etc/group</filename>).  Those user accounts should not
    be used for any other purpose!</para>

    <para>Nix will never run two builds under the same user account at
    the same time.  This is to prevent an obvious security hole: a
    malicious user writing a Nix expression that modifies the build
    result of a legitimate Nix expression being built by another user.
    Therefore it is good to have as many Nix build user accounts as
    you can spare.  (Remember: uids are cheap.)</para>

    <para>The build users should have permission to create files in
    the Nix store, but not delete them.  Therefore,
    <filename>/nix/store</filename> should be owned by the Nix
    account, its group should be the group specified here, and its
    mode should be <literal>1775</literal>.</para>

    <para>If the build users group is empty, builds will be performed
    under the uid of the Nix process (that is, the uid of the caller
    if <envar>NIX_REMOTE</envar> is empty, the uid under which the Nix
    daemon runs if <envar>NIX_REMOTE</envar> is
    <literal>daemon</literal>).  Obviously, this should not be used in
    multi-user settings with untrusted users.</para>

    </listitem>

  </varlistentry>


  <varlistentry><term><literal>sandbox</literal></term>

    <listitem><para>If set to <literal>true</literal>, builds will be
    performed in a <emphasis>sandboxed environment</emphasis>, i.e.,
    they’re isolated from the normal file system hierarchy and will
    only see their dependencies in the Nix store, the temporary build
    directory, private versions of <filename>/proc</filename>,
    <filename>/dev</filename>, <filename>/dev/shm</filename> and
    <filename>/dev/pts</filename> (on Linux), and the paths configured with the
    <link linkend='conf-sandbox-paths'><literal>sandbox-paths</literal>
    option</link>. This is useful to prevent undeclared dependencies
    on files in directories such as <filename>/usr/bin</filename>. In
    addition, on Linux, builds run in private PID, mount, network, IPC
    and UTS namespaces to isolate them from other processes in the
    system (except that fixed-output derivations do not run in private
    network namespace to ensure they can access the network).</para>

    <para>Currently, sandboxing only work on Linux and macOS. The use
    of a sandbox requires that Nix is run as root (so you should use
    the <link linkend='conf-build-users-group'>“build users”
    feature</link> to perform the actual builds under different users
    than root).</para>

    <para>If this option is set to <literal>relaxed</literal>, then
    fixed-output derivations and derivations that have the
    <varname>__noChroot</varname> attribute set to
    <literal>true</literal> do not run in sandboxes.</para>

    <para>The default is <literal>false</literal>.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-sandbox-paths">
    <term><literal>sandbox-paths</literal></term>

    <listitem><para>A list of paths bind-mounted into Nix sandbox
    environments. You can use the syntax
    <literal><replaceable>target</replaceable>=<replaceable>source</replaceable></literal>
    to mount a path in a different location in the sandbox; for
    instance, <literal>/bin=/nix-bin</literal> will mount the path
    <literal>/nix-bin</literal> as <literal>/bin</literal> inside the
    sandbox. If <replaceable>source</replaceable> is followed by
    <literal>?</literal>, then it is not an error if
    <replaceable>source</replaceable> does not exist; for example,
    <literal>/dev/nvidiactl?</literal> specifies that
    <filename>/dev/nvidiactl</filename> will only be mounted in the
    sandbox if it exists in the host filesystem.</para>

    <para>Depending on how Nix was built, the default value for this option
    may be empty or provide <filename>/bin/sh</filename> as a
    bind-mount of <command>bash</command>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-extra-sandbox-paths">
    <term><literal>build-extra-sandbox-paths</literal></term>

    <listitem><para>A list of additional paths appended to
    <option>sandbox-paths</option>. Useful if you want to extend
    its default value.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>use-substitutes</literal></term>

    <listitem><para>If set to <literal>true</literal> (default), Nix
    will use binary substitutes if available.  This option can be
    disabled to force building from source.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>fallback</literal></term>

    <listitem><para>If set to <literal>true</literal>, Nix will fall
    back to building from source if a binary substitute fails.  This
    is equivalent to the <option>--fallback</option> flag.  The
    default is <literal>false</literal>.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>keep-build-log</literal></term>

    <listitem><para>If set to <literal>true</literal> (the default),
    Nix will write the build log of a derivation (i.e. the standard
    output and error of its builder) to the directory
    <filename>/nix/var/log/nix/drvs</filename>.  The build log can be
    retrieved using the command <command>nix-store -l
    <replaceable>path</replaceable></command>.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>compress-build-log</literal></term>

    <listitem><para>If set to <literal>true</literal> (the default),
    build logs written to <filename>/nix/var/log/nix/drvs</filename>
    will be compressed on the fly using bzip2.  Otherwise, they will
    not be compressed.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>substituters</literal></term>

    <listitem><para>A list of URLs of substituters, separated by
    whitespace.  The default is
    <literal>https://cache.nixos.org</literal>.</para></listitem>

  </varlistentry>


  <!--
  <varlistentry><term><literal>binary-caches-files</literal></term>

    <listitem><para>A list of names of files that will be read to
    obtain additional binary cache URLs.  The default is
    <literal>/nix/var/nix/profiles/per-user/<replaceable>username</replaceable>/channels/binary-caches/*</literal>.
    Note that when you’re using the Nix daemon,
    <replaceable>username</replaceable> is always equal to
    <literal>root</literal>, so Nix will only use the binary caches
    provided by the channels installed by root.  Do not set this
    option to read files created by untrusted users!</para></listitem>

  </varlistentry>
  -->


  <varlistentry><term><literal>trusted-substituters</literal></term>

    <listitem><para>A list of URLs of substituters, separated by
    whitespace.  These are not used by default, but can be enabled by
    users of the Nix daemon by specifying <literal>--option
    substituters <replaceable>urls</replaceable></literal> on the
    command line.  Unprivileged users are only allowed to pass a
    subset of the URLs listed in <literal>substituters</literal> and
    <literal>trusted-substituters</literal>.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>extra-substituters</literal></term>

    <listitem><para>Additional binary caches appended to those
    specified in <option>substituters</option>.  When used by
    unprivileged users, untrusted substituters (i.e. those not listed
    in <option>trusted-substituters</option>) are silently
    ignored.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>signed-binary-caches</literal></term>

    <listitem><para>If set to <literal>*</literal> (the default), Nix
    will only download binaries if they are signed using one of the
    keys listed in <option>trusted-public-keys</option>. Set to
    the empty string to disable signature checking.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>trusted-public-keys</literal></term>

    <listitem><para>A whitespace-separated list of public keys. When
    paths are copied from another Nix store (such as a binary cache),
    they must be signed with one of these keys. For example:
    <literal>cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY=
    hydra.nixos.org-1:CNHJZBh9K4tP3EKF6FkkgeVYsS3ohTl+oS0Qa8bezVs=</literal>.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>http-connections</literal></term>

    <listitem><para>The maximum number of parallel TCP connections
    used to fetch files from binary caches and by other downloads. It
    defaults to 25. 0 means no limit.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>netrc-file</literal></term>

    <listitem><para>If set to an absolute path to a <filename>netrc</filename>
    file, Nix will use the HTTP authentication credentials in this file when
    trying to download from a remote host through HTTP or HTTPS. Defaults to
    <filename>$NIX_CONF_DIR/netrc</filename>.</para>

    <para>The <filename>netrc</filename> file consists of a list of
    accounts in the following format:

<screen>
machine <replaceable>my-machine</replaceable>
login <replaceable>my-username</replaceable>
password <replaceable>my-password</replaceable>
</screen>

    For the exact syntax, see <link
    xlink:href="https://ec.haxx.se/usingcurl-netrc.html">the
    <literal>curl</literal> documentation.</link></para></listitem>

  </varlistentry>


  <varlistentry><term><literal>system</literal></term>

    <listitem><para>This option specifies the canonical Nix system
    name of the current installation, such as
    <literal>i686-linux</literal> or
    <literal>x86_64-darwin</literal>.  Nix can only build derivations
    whose <literal>system</literal> attribute equals the value
    specified here.  In general, it never makes sense to modify this
    value from its default, since you can use it to ‘lie’ about the
    platform you are building on (e.g., perform a Mac OS build on a
    Linux machine; the result would obviously be wrong).  It only
    makes sense if the Nix binaries can run on multiple platforms,
    e.g., ‘universal binaries’ that run on <literal>x86_64-linux</literal> and
    <literal>i686-linux</literal>.</para>

    <para>It defaults to the canonical Nix system name detected by
    <filename>configure</filename> at build time.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>fsync-metadata</literal></term>

    <listitem><para>If set to <literal>true</literal>, changes to the
    Nix store metadata (in <filename>/nix/var/nix/db</filename>) are
    synchronously flushed to disk.  This improves robustness in case
    of system crashes, but reduces performance.  The default is
    <literal>true</literal>.</para></listitem>

  </varlistentry>


  <varlistentry><term><literal>auto-optimise-store</literal></term>

    <listitem><para>If set to <literal>true</literal>, Nix
    automatically detects files in the store that have identical
    contents, and replaces them with hard links to a single copy.
    This saves disk space.  If set to <literal>false</literal> (the
    default), you can still run <command>nix-store
    --optimise</command> to get rid of duplicate
    files.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-connect-timeout"><term><literal>connect-timeout</literal></term>

    <listitem>

      <para>The timeout (in seconds) for establishing connections in
      the binary cache substituter.  It corresponds to
      <command>curl</command>’s <option>--connect-timeout</option>
      option.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-trusted-users"><term><literal>trusted-users</literal></term>

    <listitem>

      <para>A list of names of users (separated by whitespace) that
      have additional rights when connecting to the Nix daemon, such
      as the ability to specify additional binary caches, or to import
      unsigned NARs. You can also specify groups by prefixing them
      with <literal>@</literal>; for instance,
      <literal>@wheel</literal> means all users in the
      <literal>wheel</literal> group. The default is
      <literal>root</literal>.</para>

      <warning><para>The users listed here have the ability to
      compromise the security of a multi-user Nix store. For instance,
      they could install Trojan horses subsequently executed by other
      users. So you should consider carefully whether to add users to
      this list.</para></warning>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allowed-users"><term><literal>allowed-users</literal></term>

    <listitem>

      <para>A list of names of users (separated by whitespace) that
      are allowed to connect to the Nix daemon. As with the
      <option>trusted-users</option> option, you can specify groups by
      prefixing them with <literal>@</literal>. Also, you can allow
      all users by specifying <literal>*</literal>. The default is
      <literal>*</literal>.</para>

      <para>Note that trusted users are always allowed to connect.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-restrict-eval"><term><literal>restrict-eval</literal></term>

    <listitem>

      <para>If set to <literal>true</literal>, the Nix evaluator will
      not allow access to any files outside of the Nix search path (as
      set via the <envar>NIX_PATH</envar> environment variable or the
      <option>-I</option> option), or to URIs outside of
      <option>allowed-uri</option>. The default is
      <literal>false</literal>.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allowed-uris"><term><literal>allowed-uris</literal></term>

    <listitem>

      <para>A list of URI prefixes to which access is allowed in
      restricted evaluation mode. For example, when set to
      <literal>https://github.com/NixOS</literal>, builtin functions
      such as <function>fetchGit</function> are allowed to access
      <literal>https://github.com/NixOS/patchelf.git</literal>.</para>

    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-pre-build-hook"><term><literal>pre-build-hook</literal></term>

    <listitem>


      <para>If set, the path to a program that can set extra
      derivation-specific settings for this system. This is used for settings
      that can't be captured by the derivation model itself and are too variable
      between different versions of the same system to be hard-coded into nix.
      </para>

      <para>The hook is passed the derivation path and, if sandboxes are enabled,
      the sandbox directory. It can then modify the sandbox and send a series of
      commands to modify various settings to stdout. The currently recognized
      commands are:</para>

      <variablelist>
        <varlistentry xml:id="extra-sandbox-paths">
          <term><literal>extra-sandbox-paths</literal></term>

          <listitem>

            <para>Pass a list of files and directories to be included in the
            sandbox for this build. One entry per line, terminated by an empty
            line. Entries have the same format as
            <literal>sandbox-paths</literal>.</para>

          </listitem>

        </varlistentry>
      </variablelist>
    </listitem>

  </varlistentry>


  <varlistentry xml:id="conf-repeat"><term><literal>repeat</literal></term>

    <listitem><para>How many times to repeat builds to check whether
    they are deterministic. The default value is 0. If the value is
    non-zero, every build is repeated the specified number of
    times. If the contents of any of the runs differs from the
    previous ones, the build is rejected and the resulting store paths
    are not registered as “valid” in Nix’s database.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-sandbox-dev-shm-size"><term><literal>sandbox-dev-shm-size</literal></term>

    <listitem><para>This option determines the maximum size of the
    <literal>tmpfs</literal> filesystem mounted on
    <filename>/dev/shm</filename> in Linux sandboxes. For the format,
    see the description of the <option>size</option> option of
    <literal>tmpfs</literal> in
    <citerefentry><refentrytitle>mount</refentrytitle><manvolnum>8</manvolnum></citerefentry>. The
    default is <literal>50%</literal>.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allow-import-from-derivation"><term><literal>allow-import-from-derivation</literal></term>

    <listitem><para>By default, Nix allows you to <function>import</function> from a derivation,
    allowing building at evaluation time. With this option set to false, Nix will throw an error
    when evaluating an expression that uses this feature, allowing users to ensure their evaluation
    will not require any builds to take place.</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-allow-new-privileges"><term><literal>allow-new-privileges</literal></term>

    <listitem><para>(Linux-specific.) By default, builders on Linux
    cannot acquire new privileges by calling setuid/setgid programs or
    programs that have file capabilities. For example, programs such
    as <command>sudo</command> or <command>ping</command> will
    fail. (Note that in sandbox builds, no such programs are available
    unless you bind-mount them into the sandbox via the
    <option>sandbox-paths</option> option.) You can allow the
    use of such programs by enabling this option. This is impure and
    usually undesirable, but may be useful in certain scenarios
    (e.g. to spin up containers or set up userspace network interfaces
    in tests).</para></listitem>

  </varlistentry>


  <varlistentry xml:id="conf-hashed-mirrors"><term><literal>hashed-mirrors</literal></term>

    <listitem><para>A list of web servers used by
    <function>builtins.fetchurl</function> to obtain files by
    hash. The default is
    <literal>http://tarballs.nixos.org/</literal>. Given a hash type
    <replaceable>ht</replaceable> and a base-16 hash
    <replaceable>h</replaceable>, Nix will try to download the file
    from
    <literal>hashed-mirror/<replaceable>ht</replaceable>/<replaceable>h</replaceable></literal>.
    This allows files to be downloaded even if they have disappeared
    from their original URI. For example, given the default mirror
    <literal>http://tarballs.nixos.org/</literal>, when building the derivation

<programlisting>
builtins.fetchurl {
  url = https://example.org/foo-1.2.3.tar.xz;
  sha256 = "2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae";
}
</programlisting>

    Nix will attempt to download this file from
    <literal>http://tarballs.nixos.org/sha256/2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae</literal>
    first. If it is not available there, if will try the original URI.</para></listitem>

  </varlistentry>


</variablelist>

</para>

</refsection>

</refentry>