chiark / gitweb /
man: document the slice and scope units, add systemd.cgroup(5)
authorZbigniew Jędrzejewski-Szmek <zbyszek@in.waw.pl>
Wed, 10 Jul 2013 05:25:02 +0000 (01:25 -0400)
committerZbigniew Jędrzejewski-Szmek <zbyszek@in.waw.pl>
Fri, 12 Jul 2013 05:10:04 +0000 (01:10 -0400)
Makefile-man.am
man/systemd.cgroup.xml [new file with mode: 0644]
man/systemd.exec.xml
man/systemd.mount.xml
man/systemd.scope.xml [new file with mode: 0644]
man/systemd.service.xml
man/systemd.slice.xml [new file with mode: 0644]
man/systemd.socket.xml
man/systemd.swap.xml

index bef749a..e74ccea 100644 (file)
@@ -76,6 +76,7 @@ MANPAGES += \
        man/systemd-update-utmp.service.8 \
        man/systemd.1 \
        man/systemd.automount.5 \
+       man/systemd.cgroup.5 \
        man/systemd.device.5 \
        man/systemd.exec.5 \
        man/systemd.journal-fields.7 \
@@ -83,7 +84,9 @@ MANPAGES += \
        man/systemd.mount.5 \
        man/systemd.path.5 \
        man/systemd.preset.5 \
+       man/systemd.scope.5 \
        man/systemd.service.5 \
+       man/systemd.slice.5 \
        man/systemd.snapshot.5 \
        man/systemd.socket.5 \
        man/systemd.special.7 \
diff --git a/man/systemd.cgroup.xml b/man/systemd.cgroup.xml
new file mode 100644 (file)
index 0000000..504c968
--- /dev/null
@@ -0,0 +1,308 @@
+<?xml version='1.0'?> <!--*-nxml-*-->
+<?xml-stylesheet type="text/xsl" href="http://docbook.sourceforge.net/release/xsl/current/xhtml/docbook.xsl"?>
+<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
+"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
+
+<!--
+This file is part of systemd.
+
+Copyright 2013 Zbigniew Jędrzejewski-Szmek
+
+systemd is free software; you can redistribute it and/or modify it
+under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation; either version 2.1 of the License, or
+(at your option) any later version.
+
+systemd is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+Lesser General Public License for more details.
+
+You should have received a copy of the GNU Lesser General Public License
+along with systemd; If not, see <http://www.gnu.org/licenses/>.
+-->
+
+<refentry id="systemd.cgroup">
+  <refentryinfo>
+    <title>systemd.cgroup</title>
+    <productname>systemd</productname>
+
+    <authorgroup>
+      <author>
+        <contrib>Developer</contrib>
+        <firstname>Lennart</firstname>
+        <surname>Poettering</surname>
+        <email>lennart@poettering.net</email>
+      </author>
+    </authorgroup>
+  </refentryinfo>
+
+  <refmeta>
+    <refentrytitle>systemd.cgroup</refentrytitle>
+    <manvolnum>5</manvolnum>
+  </refmeta>
+
+  <refnamediv>
+    <refname>systemd.cgroup</refname>
+    <refpurpose>Cgroup configuration unit settings</refpurpose>
+  </refnamediv>
+
+  <refsynopsisdiv>
+    <para>
+      <filename><replaceable>slice</replaceable>.slice</filename>,
+      <filename><replaceable>scope</replaceable>.scope</filename>,
+      <filename><replaceable>service</replaceable>.service</filename>,
+      <filename><replaceable>socket</replaceable>.socket</filename>,
+      <filename><replaceable>mount</replaceable>.mount</filename>,
+      <filename><replaceable>swap</replaceable>.swap</filename>
+    </para>
+  </refsynopsisdiv>
+
+  <refsect1>
+    <title>Description</title>
+
+    <para>Unit configuration files for services, slices, scopes,
+    sockets, mount points, and swap devices share a subset of
+    configuration options which configure the control group settings
+    for spawned processes.</para>
+
+    <para>This man page lists the configuration options shared by
+    those six unit types. See
+    <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+    for the common options of all unit configuration files, and
+    <citerefentry><refentrytitle>systemd.slice</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+    <citerefentry><refentrytitle>systemd.scope</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+    <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+    <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+    <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+    and
+    <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+    for more information on the specific unit configuration files. The
+    execution specific configuration options are configured in the
+    [Slice], [Scope], [Service], [Socket], [Mount], or [Swap]
+    sections, depending on the unit type.</para>
+  </refsect1>
+
+  <refsect1>
+    <title>Options</title>
+
+    <para>Units of the types listed above can have settings
+    for cgroup configuration:</para>
+
+    <variablelist class='unit-directives'>
+      <varlistentry>
+        <term><varname>CPUAccounting=</varname></term>
+
+        <listitem>
+          <para>Turn on the CPU usage accounting for this
+          unit.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>BlockIOAccounting=</varname></term>
+
+        <listitem>
+          <para>Turn on the Block IO bandwidth accounting
+          for this unit.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>MemoryAccounting=</varname></term>
+
+        <listitem>
+          <para>Turn on the process and kernel memory
+          accounting for this unit.</para>
+        </listitem>
+      </varlistentry>
+
+
+      <varlistentry>
+        <term><varname>CPUShares=<replaceable>weight</replaceable></varname></term>
+
+        <listitem>
+          <para>Assign the specified overall CPU time share weight to
+          the processes executed. Takes an integer value. This
+          controls the <literal>cpu.shares</literal> control group
+          attribute, which defaults to 1024. For details about this
+          control group attribute see <ulink
+          url="http://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt">sched-design-CFS.txt</ulink>.</para>
+
+          <para>Implies <literal>CPUAccounting=true</literal>.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>MemoryLimit=<replaceable>bytes</replaceable></varname></term>
+        <term><varname>MemorySoftLimit=<replaceable>bytes</replaceable></varname></term>
+
+        <listitem>
+          <para>Specify the hard and soft limits on maximum memory
+          usage of the executed processes. The "hard" limit specifies
+          how much process and kernel memory can be used by tasks in
+          this unit, when there is no memory contention. If the kernel
+          detects memory contention, memory reclaim will be performed
+          until the memory usage is within the "soft" limit. Takes a
+          memory size in bytes. If the value is suffixed with K, M, G
+          or T the specified memory size is parsed as Kilobytes,
+          Megabytes, Gigabytes, or Terabytes (with the base 1024),
+          respectively. This controls the
+          <literal>memory.limit_in_bytes</literal> and
+          <literal>memory.soft_limit_in_bytes</literal> control group
+          attributes. For details about these control group attributes
+          see <ulink
+          url="http://www.kernel.org/doc/Documentation/cgroups/memory.txt">memory.txt</ulink>.</para>
+
+          <para>Implies <literal>MemoryAccounting=true</literal>.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>BlockIOWeight=<replaceable>weight</replaceable></varname></term>
+
+        <listitem><para>Set the default
+        overall block IO weight for the
+        executed processes. Takes a single
+        weight value (between 10 and 1000) to
+        set the default block IO weight. This
+        controls the
+        <literal>blkio.weight</literal>
+        control group attribute, which
+        defaults to 1000. For details about
+        this control group attribute see
+        <ulink
+            url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.</para></listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>BlockIODeviceWeight=<replaceable>device</replaceable> <replaceable>weight</replaceable></varname></term>
+
+        <listitem>
+          <para>Set the per-device overall block IO weight for the
+          executed processes. Takes a space-separated pair of a file
+          path and a weight value to specify the device specific
+          weight value, between 10 and 1000. (Example: "/dev/sda
+          500"). The file path may be specified as path to a block
+          device node or as any other file in which case the backing
+          block device of the file system of the file is
+          determined. This controls the
+          <literal>blkio.weight_device</literal> control group
+          attribute, which defaults to 1000. Use this option multiple
+          times to set weights for multiple devices. For details about
+          this control group attribute see <ulink
+          url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>BlockIOReadBandwidth=<replaceable>device</replaceable> <replaceable>bytes</replaceable></varname></term>
+        <term><varname>BlockIOWriteBandwidth=<replaceable>device</replaceable> <replaceable>bytes</replaceable></varname></term>
+
+        <listitem>
+          <para>Set the per-device overall block IO bandwidth limit
+          for the executed processes. Takes a space-separated pair of
+          a file path and a bandwidth value (in bytes per second) to
+          specify the device specific bandwidth. The file path may be
+          a path to a block device node, or as any other file in which
+          case the backing block device of the file system of the file
+          is used.  If the bandwidth is suffixed with K, M, G, or T
+          the specified bandwidth is parsed as Kilobytes, Megabytes,
+          Gigabytes, or Terabytes, respectively (Example:
+          "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 5M"). This
+          controls the <literal>blkio.read_bps_device</literal> and
+          <literal>blkio.write_bps_device</literal> control group
+          attributes. Use this option multiple times to set bandwidth
+          limits for multiple devices. For details about these control
+          group attributes see
+          <ulink url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.
+          </para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>DeviceAllow=</varname></term>
+
+        <listitem>
+          <para>Control access to specific device nodes by the
+          executed processes. Takes two space-separated strings: a
+          device node path (such as <filename>/dev/null</filename>)
+          followed by a combination of <constant>r</constant>,
+          <constant>w</constant>, <constant>m</constant> to control
+          <emphasis>r</emphasis>eading, <emphasis>w</emphasis>riting,
+          or creating of the specific device node by the unit
+          (<emphasis>m</emphasis>knod), respectively. This controls
+          the <literal>devices.allow</literal> and
+          <literal>devices.deny</literal> control group
+          attributes. For details about these control group attributes
+          see <ulink
+          url="http://www.kernel.org/doc/Documentation/cgroups/devices.txt">devices.txt</ulink>.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><varname>DevicePolicy=auto|closed|strict</varname></term>
+
+        <listitem>
+          <para>
+            Control the policy for allowing device access:
+          </para>
+          <variablelist>
+            <varlistentry>
+              <term><option>strict</option></term>
+              <listitem>
+                <para>means to only allow types of access that are
+                explicitly specified.</para>
+              </listitem>
+            </varlistentry>
+
+            <varlistentry>
+              <term><option>closed</option></term>
+              <listitem>
+                <para>in addition allows access to standard pseudo
+                devices including
+                <filename>/dev/null</filename>,
+                <filename>/dev/zero</filename>,
+                <filename>/dev/full</filename>,
+                <filename>/dev/random</filename>, and
+                <filename>/dev/urandom</filename>.
+                </para>
+              </listitem>
+            </varlistentry>
+
+            <varlistentry>
+              <term><option>auto</option></term>
+              <listitem>
+                <para>
+                  in addition allows access to all devices if no
+                  explicit <varname>DeviceAllow=</varname> is present.
+                  This is the default.
+                </para>
+              </listitem>
+            </varlistentry>
+          </variablelist>
+        </listitem>
+      </varlistentry>
+    </variablelist>
+  </refsect1>
+
+  <refsect1>
+    <title>See Also</title>
+    <para>
+      <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.slice</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.scope</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
+      The documentation for control groups and specific controllers in the Linux kernel:
+      <ulink url="http://www.kernel.org/doc/Documentation/cgroups/cgroups.txt">cgroups.txt</ulink>,
+      <ulink url="http://www.kernel.org/doc/Documentation/cgroups/cpuacct.txt">cpuacct.txt</ulink>,
+      <ulink url="http://www.kernel.org/doc/Documentation/cgroups/memory.txt">memory.txt</ulink>,
+      <ulink url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.
+    </para>
+  </refsect1>
+</refentry>
index 4294e54..d299fc0 100644 (file)
@@ -58,7 +58,7 @@
                 <title>Description</title>
 
                 <para>Unit configuration files for services, sockets,
-                mount points and swap devices share a subset of
+                mount points, and swap devices share a subset of
                 configuration options which define the execution
                 environment of spawned processes.</para>
 
@@ -69,7 +69,7 @@
                 files, and
                 <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 <citerefentry><refentrytitle>systemd.socket</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
-                <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+                <citerefentry><refentrytitle>systemd.swap</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 and
                 <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>
                 for more information on the specific unit
                         </varlistentry>
 
                         <varlistentry>
-                                <term><varname>CPUShares=</varname></term>
-
-                                <listitem><para>Assign the specified
-                                overall CPU time shares to the
-                                processes executed. Takes an integer
-                                value. This controls the
-                                <literal>cpu.shares</literal> control
-                                group attribute, which defaults to
-                                1024. For details about this control
-                                group attribute see <ulink
-                                url="http://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt">sched-design-CFS.txt</ulink>.</para></listitem>
-                        </varlistentry>
-
-                        <varlistentry>
-                                <term><varname>MemoryLimit=</varname></term>
-                                <term><varname>MemorySoftLimit=</varname></term>
-
-                                <listitem><para>Limit the overall memory usage
-                                of the executed processes to a certain
-                                size. Takes a memory size in bytes. If
-                                the value is suffixed with K, M, G or
-                                T the specified memory size is parsed
-                                as Kilobytes, Megabytes, Gigabytes,
-                                or Terabytes (to the base
-                                1024), respectively. This controls the
-                                <literal>memory.limit_in_bytes</literal>
-                                and
-                                <literal>memory.soft_limit_in_bytes</literal>
-                                control group attributes. For details
-                                about these control group attributes
-                                see <ulink
-                                url="http://www.kernel.org/doc/Documentation/cgroups/memory.txt">memory.txt</ulink>.</para></listitem>
-                        </varlistentry>
-
-                        <varlistentry>
-                                <term><varname>DeviceAllow=</varname></term>
-                                <term><varname>DeviceDeny=</varname></term>
-
-                                <listitem><para>Control access to
-                                specific device nodes by the executed processes. Takes two
-                                space-separated strings: a device node
-                                path (such as
-                                <filename>/dev/null</filename>)
-                                followed by a combination of r, w, m
-                                to control reading, writing, or
-                                creating of the specific device node
-                                by the unit, respectively. This controls the
-                                <literal>devices.allow</literal>
-                                and
-                                <literal>devices.deny</literal>
-                                control group attributes. For details
-                                about these control group attributes
-                                see <ulink
-                                url="http://www.kernel.org/doc/Documentation/cgroups/devices.txt">devices.txt</ulink>.</para></listitem>
-                        </varlistentry>
-
-                        <varlistentry>
-                                <term><varname>BlockIOWeight=</varname></term>
-
-                                <listitem><para>Set the default or
-                                per-device overall block IO weight
-                                value for the executed
-                                processes. Takes either a single
-                                weight value (between 10 and 1000) to
-                                set the default block IO weight, or a
-                                space-separated pair of a file path
-                                and a weight value to specify the
-                                device specific weight value (Example:
-                                "/dev/sda 500"). The file path may be
-                                specified as path to a block device
-                                node or as any other file in which
-                                case the backing block device of the
-                                file system of the file is
-                                determined. This controls the
-                                <literal>blkio.weight</literal> and
-                                <literal>blkio.weight_device</literal>
-                                control group attributes, which
-                                default to 1000. Use this option
-                                multiple times to set weights for
-                                multiple devices. For details about
-                                these control group attributes see
-                                <ulink
-                                url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.</para></listitem>
-                        </varlistentry>
-
-                        <varlistentry>
-                                <term><varname>BlockIOReadBandwidth=</varname></term>
-                                <term><varname>BlockIOWriteBandwidth=</varname></term>
-
-                                <listitem><para>Set the per-device
-                                overall block IO bandwidth limit for
-                                the executed processes. Takes a
-                                space-separated pair of a file path and a
-                                bandwidth value (in bytes per second)
-                                to specify the device specific
-                                bandwidth. The file path may be
-                                specified as path to a block device
-                                node or as any other file in which
-                                case the backing block device of the
-                                file system of the file is determined.
-                                If the bandwidth is suffixed with K, M,
-                                G, or T the specified bandwidth is
-                                parsed as Kilobytes, Megabytes,
-                                Gigabytes, or Terabytes, respectively (Example:
-                                "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0
-                                5M"). This controls the
-                                <literal>blkio.read_bps_device</literal>
-                                and
-                                <literal>blkio.write_bps_device</literal>
-                                control group attributes. Use this
-                                option multiple times to set bandwidth
-                                limits for multiple devices. For
-                                details about these control group
-                                attributes see <ulink
-                                url="http://www.kernel.org/doc/Documentation/cgroups/blkio-controller.txt">blkio-controller.txt</ulink>.</para></listitem>
-                        </varlistentry>
-
-                        <varlistentry>
                                 <term><varname>ReadWriteDirectories=</varname></term>
                                 <term><varname>ReadOnlyDirectories=</varname></term>
                                 <term><varname>InaccessibleDirectories=</varname></term>
index c57f613..214f40c 100644 (file)
                 which define the execution environment the
                 <citerefentry><refentrytitle>mount</refentrytitle><manvolnum>8</manvolnum></citerefentry>
                 binary is executed in, and in
-                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
-                which define the way the processes are
-                terminated. Note that the User= and Group= options are
-                not particularly useful for mount units specifying a
-                <literal>Type=</literal> option or using configuration
-                not specified in <filename>/etc/fstab</filename>;
+                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                which define the way the processes are terminated, and
+                in
+                <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                which configure control group settings for the
+                processes of the service. Note that the User= and
+                Group= options are not particularly useful for mount
+                units specifying a <literal>Type=</literal> option or
+                using configuration not specified in
+                <filename>/etc/fstab</filename>;
                 <citerefentry><refentrytitle>mount</refentrytitle><manvolnum>8</manvolnum></citerefentry>
                 will refuse options that aren't listed in
                 <filename>/etc/fstab</filename> if it is not run as
                           <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                          <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.device</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>proc</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
diff --git a/man/systemd.scope.xml b/man/systemd.scope.xml
new file mode 100644 (file)
index 0000000..31f2d6f
--- /dev/null
@@ -0,0 +1,100 @@
+<?xml version='1.0'?> <!--*-nxml-*-->
+<?xml-stylesheet type="text/xsl" href="http://docbook.sourceforge.net/release/xsl/current/xhtml/docbook.xsl"?>
+<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
+"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
+
+<!--
+This file is part of systemd.
+
+Copyright 2013 Zbigniew Jędrzejewski-Szmek
+
+systemd is free software; you can redistribute it and/or modify it
+under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation; either version 2.1 of the License, or
+(at your option) any later version.
+
+systemd is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+Lesser General Public License for more details.
+
+You should have received a copy of the GNU Lesser General Public License
+along with systemd; If not, see <http://www.gnu.org/licenses/>.
+-->
+
+<refentry id="systemd.scope">
+  <refentryinfo>
+    <title>systemd.scope</title>
+    <productname>systemd</productname>
+
+    <authorgroup>
+      <author>
+        <contrib>Developer</contrib>
+        <firstname>Lennart</firstname>
+        <surname>Poettering</surname>
+        <email>lennart@poettering.net</email>
+      </author>
+    </authorgroup>
+  </refentryinfo>
+
+  <refmeta>
+    <refentrytitle>systemd.scope</refentrytitle>
+    <manvolnum>5</manvolnum>
+  </refmeta>
+
+  <refnamediv>
+    <refname>systemd.scope</refname>
+    <refpurpose>Scope unit configuration</refpurpose>
+  </refnamediv>
+
+  <refsynopsisdiv>
+    <para><filename><replaceable>scope</replaceable>.scope</filename></para>
+  </refsynopsisdiv>
+
+  <refsect1>
+    <title>Description</title>
+
+    <para>A unit configuration file whose name ends in
+    <literal>.scope</literal> encodes information about a unit created
+    by systemd to encapsulate processes launched not by systemd
+    itself. This management is performed by creating a node in the
+    control group tree. Processes are moved into the scope by means
+    of the DBus API.
+    <command>systemd-run <option>--scope</option></command> can be
+    used to easily launch a command in a new scope unit.</para>
+
+    <para>See
+    <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+    for the common options of all unit configuration
+    files. The common configuration items are configured
+    in the generic [Unit] and [Install] sections. The
+    scope specific configuration options are configured in
+    the [Scope] section. Currently, only generic cgroup settings
+    as described in
+    <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>7</manvolnum></citerefentry> are allowed.
+    </para>
+
+    <para>Unless <varname>DefaultDependencies=false</varname>
+    is used, scope units will implicitly have dependencies of
+    type <varname>Conflicts=</varname> and
+    <varname>Before=</varname> on
+    <filename>shutdown.target</filename>. These ensure
+    that scope units are removed prior to system
+    shutdown. Only scope units involved with early boot or
+    late system shutdown should disable this option.
+    </para>
+  </refsect1>
+
+  <refsect1>
+    <title>See Also</title>
+    <para>
+      <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd-run</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
+    </para>
+  </refsect1>
+
+</refentry>
index abe3a8d..238a49a 100644 (file)
                 <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 which define the execution environment the commands
                 are executed in, and in
-                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 which define the way the processes of the service are
-                terminated.</para>
+                terminated, and in
+                <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                which configure control group settings for the
+                processes of the service.</para>
 
                 <para>Unless <varname>DefaultDependencies=</varname>
                 is set to <option>false</option>, service units will
                           <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                          <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>
                   </para>
diff --git a/man/systemd.slice.xml b/man/systemd.slice.xml
new file mode 100644 (file)
index 0000000..5376921
--- /dev/null
@@ -0,0 +1,99 @@
+<?xml version='1.0'?> <!--*-nxml-*-->
+<?xml-stylesheet type="text/xsl" href="http://docbook.sourceforge.net/release/xsl/current/xhtml/docbook.xsl"?>
+<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
+"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
+
+<!--
+This file is part of systemd.
+
+Copyright 2013 Zbigniew Jędrzejewski-Szmek
+
+systemd is free software; you can redistribute it and/or modify it
+under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation; either version 2.1 of the License, or
+(at your option) any later version.
+
+systemd is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+Lesser General Public License for more details.
+
+You should have received a copy of the GNU Lesser General Public License
+along with systemd; If not, see <http://www.gnu.org/licenses/>.
+-->
+
+<refentry id="systemd.slice">
+  <refentryinfo>
+    <title>systemd.slice</title>
+    <productname>systemd</productname>
+
+    <authorgroup>
+      <author>
+        <contrib>Developer</contrib>
+        <firstname>Lennart</firstname>
+        <surname>Poettering</surname>
+        <email>lennart@poettering.net</email>
+      </author>
+    </authorgroup>
+  </refentryinfo>
+
+  <refmeta>
+    <refentrytitle>systemd.slice</refentrytitle>
+    <manvolnum>5</manvolnum>
+  </refmeta>
+
+  <refnamediv>
+    <refname>systemd.slice</refname>
+    <refpurpose>Slice unit configuration</refpurpose>
+  </refnamediv>
+
+  <refsynopsisdiv>
+    <para><filename><replaceable>slice</replaceable>.slice</filename></para>
+  </refsynopsisdiv>
+
+  <refsect1>
+    <title>Description</title>
+
+    <para>A unit configuration file whose name ends in
+    <literal>.slice</literal> encodes information about a slice
+    created by systemd to manage resources used by a certain group of
+    processes. This management is performed by creating a node in the
+    control group tree. Those processes are part of different units,
+    usually <literal>.service</literal> units (see
+    <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>).
+    </para>
+
+    <para>See
+    <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+    for the common options of all unit configuration
+    files. The common configuration items are configured
+    in the generic [Unit] and [Install] sections. The
+    slice specific configuration options are configured in
+    the [Slice] section. Currently, only generic cgroup settings
+    as described in
+    <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>7</manvolnum></citerefentry> are allowed.
+    </para>
+
+    <para>Unless <varname>DefaultDependencies=false</varname>
+    is used, slice units will implicitly have dependencies of
+    type <varname>Conflicts=</varname> and
+    <varname>Before=</varname> on
+    <filename>shutdown.target</filename>. These ensure
+    that slice units are removed prior to system
+    shutdown. Only slice units involved with early boot or
+    late system shutdown should disable this option.
+    </para>
+  </refsect1>
+
+  <refsect1>
+    <title>See Also</title>
+    <para>
+      <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+      <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
+    </para>
+  </refsect1>
+
+</refentry>
index 1b899c4..fad5c2b 100644 (file)
                 <option>ExecStopPre=</option> and
                 <option>ExecStopPost=</option> commands are executed
                 in, and in
-                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 which define the way the processes are
-                terminated.</para>
+                terminated, and in
+                <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                which configure control group settings for the
+                processes of the service.</para>
 
                 <para>For each socket file a matching service file
                 (see
                           <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                          <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>
                   </para>
index d931c46..5e339ea 100644 (file)
                 which define the execution environment the
                 <citerefentry><refentrytitle>swapon</refentrytitle><manvolnum>8</manvolnum></citerefentry>
                 binary is executed in, and in
-                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+                <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                 which define the way the processes are
-                terminated.</para>
+                terminated, and in
+                <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                which configure control group settings for the
+                processes of the service.</para>
 
                 <para>Swap units must be named after the devices
                 or files they control. Example: the swap device
                           <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
+                          <citerefentry><refentrytitle>systemd.cgroup</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.device</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>systemd.mount</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
                           <citerefentry><refentrytitle>swapon</refentrytitle><manvolnum>8</manvolnum></citerefentry>,