man/systemd-nspawn.xml

   1 <?xml version='1.0'?> <!--*-nxml-*-->
   2 <!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
   3         "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
   4
   5 <!--
   6   This file is part of systemd.
   7
   8   Copyright 2010 Lennart Poettering
   9
  10   systemd is free software; you can redistribute it and/or modify it
  11   under the terms of the GNU Lesser General Public License as published by
  12   the Free Software Foundation; either version 2.1 of the License, or
  13   (at your option) any later version.
  14
  15   systemd is distributed in the hope that it will be useful, but
  16   WITHOUT ANY WARRANTY; without even the implied warranty of
  17   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  18   Lesser General Public License for more details.
  19
  20   You should have received a copy of the GNU Lesser General Public License
  21   along with systemd; If not, see <http://www.gnu.org/licenses/>.
  22 -->
  23
  24 <refentry id="systemd-nspawn">
  25
  26         <refentryinfo>
  27                 <title>systemd-nspawn</title>
  28                 <productname>systemd</productname>
  29
  30                 <authorgroup>
  31                         <author>
  32                                 <contrib>Developer</contrib>
  33                                 <firstname>Lennart</firstname>
  34                                 <surname>Poettering</surname>
  35                                 <email>lennart@poettering.net</email>
  36                         </author>
  37                 </authorgroup>
  38         </refentryinfo>
  39
  40         <refmeta>
  41                 <refentrytitle>systemd-nspawn</refentrytitle>
  42                 <manvolnum>1</manvolnum>
  43         </refmeta>
  44
  45         <refnamediv>
  46                 <refname>systemd-nspawn</refname>
  47                 <refpurpose>Spawn a namespace container for debugging, testing and building</refpurpose>
  48         </refnamediv>
  49
  50         <refsynopsisdiv>
  51                 <cmdsynopsis>
  52                         <command>systemd-nspawn</command>
  53                         <arg choice="opt" rep="repeat">OPTIONS</arg>
  54                         <arg choice="opt"><replaceable>COMMAND</replaceable>
  55                         <arg choice="opt" rep="repeat">ARGS</arg>
  56                         </arg>
  57                 </cmdsynopsis>
  58                 <cmdsynopsis>
  59                         <command>systemd-nspawn</command>
  60                         <arg choice="plain">-b</arg>
  61                         <arg choice="opt" rep="repeat">OPTIONS</arg>
  62                         <arg choice="opt" rep="repeat">ARGS</arg>
  63                 </cmdsynopsis>
  64         </refsynopsisdiv>
  65
  66         <refsect1>
  67                 <title>Description</title>
  68
  69                 <para><command>systemd-nspawn</command> may be used to
  70                 run a command or OS in a light-weight namespace
  71                 container. In many ways it is similar to
  72                 <citerefentry><refentrytitle>chroot</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
  73                 but more powerful since it fully virtualizes the file
  74                 system hierarchy, as well as the process tree, the
  75                 various IPC subsystems and the host and domain
  76                 name.</para>
  77
  78                 <para><command>systemd-nspawn</command> limits access
  79                 to various kernel interfaces in the container to
  80                 read-only, such as <filename>/sys</filename>,
  81                 <filename>/proc/sys</filename> or
  82                 <filename>/sys/fs/selinux</filename>. Network
  83                 interfaces and the system clock may not be changed
  84                 from within the container. Device nodes may not be
  85                 created. The host system cannot be rebooted and kernel
  86                 modules may not be loaded from within the
  87                 container.</para>
  88
  89                 <para>Note that even though these security precautions
  90                 are taken <command>systemd-nspawn</command> is not
  91                 suitable for secure container setups. Many of the
  92                 security features may be circumvented and are hence
  93                 primarily useful to avoid accidental changes to the
  94                 host system from the container. The intended use of
  95                 this program is debugging and testing as well as
  96                 building of packages, distributions and software
  97                 involved with boot and systems management.</para>
  98
  99                 <para>In contrast to
 100                 <citerefentry><refentrytitle>chroot</refentrytitle><manvolnum>1</manvolnum></citerefentry> <command>systemd-nspawn</command>
 101                 may be used to boot full Linux-based operating systems
 102                 in a container.</para>
 103
 104                 <para>Use a tool like
 105                 <citerefentry><refentrytitle>yum</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
 106                 <citerefentry><refentrytitle>debootstrap</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
 107                 or
 108                 <citerefentry><refentrytitle>pacman</refentrytitle><manvolnum>8</manvolnum></citerefentry>
 109                 to set up an OS directory tree suitable as file system
 110                 hierarchy for <command>systemd-nspawn</command>
 111                 containers.</para>
 112
 113                 <para>Note that <command>systemd-nspawn</command> will
 114                 mount file systems private to the container to
 115                 <filename>/dev</filename>,
 116                 <filename>/run</filename> and similar. These will
 117                 not be visible outside of the container, and their
 118                 contents will be lost when the container exits.</para>
 119
 120                 <para>Note that running two
 121                 <command>systemd-nspawn</command> containers from the
 122                 same directory tree will not make processes in them
 123                 see each other. The PID namespace separation of the
 124                 two containers is complete and the containers will
 125                 share very few runtime objects except for the
 126                 underlying file system. It is however possible to
 127                 enter an existing container, see
 128                 <link linkend='example-nsenter'>Example 4</link> below.
 129                 </para>
 130
 131                 <para><command>systemd-nspawn</command> implements the
 132                 <ulink
 133                 url="http://www.freedesktop.org/wiki/Software/systemd/ContainerInterface">Container
 134                 Interface</ulink> specification.</para>
 135
 136                 <para>As a safety check
 137                 <command>systemd-nspawn</command> will verify the
 138                 existence of <filename>/etc/os-release</filename> in
 139                 the container tree before starting the container (see
 140                 <citerefentry><refentrytitle>os-release</refentrytitle><manvolnum>5</manvolnum></citerefentry>). It
 141                 might be necessary to add this file to the container
 142                 tree manually if the OS of the container is too old to
 143                 contain this file out-of-the-box.</para>
 144         </refsect1>
 145
 146         <refsect1>
 147                 <title>Incompatibility with Auditing</title>
 148
 149                 <para>Note that the kernel auditing subsystem is
 150                 currently broken when used together with
 151                 containers. We hence recommend turning it off entirely
 152                 by booting with <literal>audit=0</literal> on the
 153                 kernel command line, or by turning it off at kernel
 154                 build time. If auditing is enabled in the kernel,
 155                 operating systems booted in an nspawn container might
 156                 refuse log-in attempts.</para>
 157         </refsect1>
 158
 159         <refsect1>
 160                 <title>Options</title>
 161
 162                 <para>If option <option>-b</option> is specified, the
 163                 arguments are used as arguments for the init
 164                 binary. Otherwise, <replaceable>COMMAND</replaceable>
 165                 specifies the program to launch in the container, and
 166                 the remaining arguments are used as arguments for this
 167                 program. If <option>-b</option> is not used and no
 168                 arguments are specifed, a shell is launched in the
 169                 container.</para>
 170
 171                 <para>The following options are understood:</para>
 172
 173                 <variablelist>
 174                         <varlistentry>
 175                                 <term><option>-h</option></term>
 176                                 <term><option>--help</option></term>
 177
 178                                 <listitem><para>Prints a short help
 179                                 text and exits.</para></listitem>
 180                         </varlistentry>
 181
 182                         <varlistentry>
 183                                 <term><option>--version</option></term>
 184
 185                                 <listitem><para>Prints a version string
 186                                 and exits.</para></listitem>
 187                         </varlistentry>
 188
 189                         <varlistentry>
 190                                 <term><option>-D</option></term>
 191                                 <term><option>--directory=</option></term>
 192
 193                                 <listitem><para>Directory to use as
 194                                 file system root for the namespace
 195                                 container. If omitted, the current
 196                                 directory will be
 197                                 used.</para></listitem>
 198                         </varlistentry>
 199
 200                         <varlistentry>
 201                                 <term><option>-b</option></term>
 202                                 <term><option>--boot</option></term>
 203
 204                                 <listitem><para>Automatically search
 205                                 for an init binary and invoke it
 206                                 instead of a shell or a user supplied
 207                                 program. If this option is used, arguments
 208                                 specified on the command line are used
 209                                 as arguments for the init binary.
 210                                 </para></listitem>
 211                         </varlistentry>
 212
 213                         <varlistentry>
 214                                 <term><option>-u</option></term>
 215                                 <term><option>--user=</option></term>
 216
 217                                 <listitem><para>Run the command
 218                                 under specified user, create home
 219                                 directory and cd into it. As rest
 220                                 of systemd-nspawn, this is not
 221                                 the security feature and limits
 222                                 against accidental changes only.
 223                                 </para></listitem>
 224                         </varlistentry>
 225
 226                         <varlistentry>
 227                                 <term><option>-M</option></term>
 228                                 <term><option>--machine=</option></term>
 229
 230                                 <listitem><para>Sets the machine name
 231                                 for this container. This name may be
 232                                 used to identify this container on the
 233                                 host, and is used to initialize the
 234                                 container's hostname (which the
 235                                 container can choose to override,
 236                                 however). If not specified, the last
 237                                 component of the root directory of the
 238                                 container is used.</para></listitem>
 239                         </varlistentry>
 240
 241                         <varlistentry>
 242                                 <term><option>--slice=</option></term>
 243
 244                                 <listitem><para>Make the container
 245                                 part of the specified slice, instead
 246                                 of the
 247                                 <filename>machine.slice</filename>.</para>
 248                                 </listitem>
 249                         </varlistentry>
 250
 251                         <varlistentry>
 252                                 <term><option>--uuid=</option></term>
 253
 254                                 <listitem><para>Set the specified UUID
 255                                 for the container. The init system
 256                                 will initialize
 257                                 <filename>/etc/machine-id</filename>
 258                                 from this if this file is not set yet.
 259                                 </para></listitem>
 260                         </varlistentry>
 261
 262                         <varlistentry>
 263                                 <term><option>--private-network</option></term>
 264
 265                                 <listitem><para>Turn off networking in
 266                                 the container. This makes all network
 267                                 interfaces unavailable in the
 268                                 container, with the exception of the
 269                                 loopback device.</para></listitem>
 270                         </varlistentry>
 271
 272                         <varlistentry>
 273                                 <term><option>--read-only</option></term>
 274
 275                                 <listitem><para>Mount the root file
 276                                 system read-only for the
 277                                 container.</para></listitem>
 278                         </varlistentry>
 279
 280                         <varlistentry>
 281                                 <term><option>--capability=</option></term>
 282
 283                                 <listitem><para>List one or more
 284                                 additional capabilities to grant the
 285                                 container. Takes a comma-separated
 286                                 list of capability names, see
 287                                 <citerefentry><refentrytitle>capabilities</refentrytitle><manvolnum>7</manvolnum></citerefentry>
 288                                 for more information. Note that the
 289                                 following capabilities will be granted
 290                                 in any way: CAP_CHOWN,
 291                                 CAP_DAC_OVERRIDE, CAP_DAC_READ_SEARCH,
 292                                 CAP_FOWNER, CAP_FSETID, CAP_IPC_OWNER,
 293                                 CAP_KILL, CAP_LEASE,
 294                                 CAP_LINUX_IMMUTABLE,
 295                                 CAP_NET_BIND_SERVICE,
 296                                 CAP_NET_BROADCAST, CAP_NET_RAW,
 297                                 CAP_SETGID, CAP_SETFCAP, CAP_SETPCAP,
 298                                 CAP_SETUID, CAP_SYS_ADMIN,
 299                                 CAP_SYS_CHROOT, CAP_SYS_NICE,
 300                                 CAP_SYS_PTRACE, CAP_SYS_TTY_CONFIG,
 301                                 CAP_SYS_RESOURCE, CAP_SYS_BOOT,
 302                                 CAP_AUDIT_WRITE,
 303                                 CAP_AUDIT_CONTROL.</para></listitem>
 304                         </varlistentry>
 305
 306                         <varlistentry>
 307                                 <term><option>--link-journal=</option></term>
 308
 309                                 <listitem><para>Control whether the
 310                                 container's journal shall be made
 311                                 visible to the host system. If enabled,
 312                                 allows viewing the container's journal
 313                                 files from the host (but not vice
 314                                 versa). Takes one of
 315                                 <literal>no</literal>,
 316                                 <literal>host</literal>,
 317                                 <literal>guest</literal>,
 318                                 <literal>auto</literal>. If
 319                                 <literal>no</literal>, the journal is
 320                                 not linked. If <literal>host</literal>,
 321                                 the journal files are stored on the
 322                                 host file system (beneath
 323                                 <filename>/var/log/journal/<replaceable>machine-id</replaceable></filename>)
 324                                 and the subdirectory is bind-mounted
 325                                 into the container at the same
 326                                 location. If <literal>guest</literal>,
 327                                 the journal files are stored on the
 328                                 guest file system (beneath
 329                                 <filename>/var/log/journal/<replaceable>machine-id</replaceable></filename>)
 330                                 and the subdirectory is symlinked into the host
 331                                 at the same location. If
 332                                 <literal>auto</literal> (the default),
 333                                 and the right subdirectory of
 334                                 <filename>/var/log/journal</filename>
 335                                 exists, it will be bind mounted
 336                                 into the container. If the
 337                                 subdirectory does not exist, no
 338                                 linking is performed. Effectively,
 339                                 booting a container once with
 340                                 <literal>guest</literal> or
 341                                 <literal>host</literal> will link the
 342                                 journal persistently if further on
 343                                 the default of <literal>auto</literal>
 344                                 is used.</para></listitem>
 345                         </varlistentry>
 346
 347                         <varlistentry>
 348                                 <term><option>-j</option></term>
 349
 350                                 <listitem><para>Equivalent to
 351                                 <option>--link-journal=guest</option>.</para></listitem>
 352                         </varlistentry>
 353
 354                         <varlistentry>
 355                                 <term><option>--bind=</option></term>
 356                                 <term><option>--bind-ro=</option></term>
 357
 358                                 <listitem><para>Bind mount a file or
 359                                 directory from the host into the
 360                                 container. Either takes a path
 361                                 argument -- in which case the
 362                                 specified path will be mounted from
 363                                 the host to the same path in the
 364                                 container --, or a colon-separated
 365                                 pair of paths -- in which case the
 366                                 first specified path is the source in
 367                                 the host, and the second path is the
 368                                 destination in the container. The
 369                                 <option>--bind-ro=</option> option
 370                                 creates read-only bind
 371                                 mount.</para></listitem>
 372                         </varlistentry>
 373                 </variablelist>
 374
 375         </refsect1>
 376
 377         <refsect1>
 378                 <title>Example 1</title>
 379
 380                 <programlisting># yum -y --releasever=19 --nogpg --installroot=/srv/mycontainer --disablerepo='*' --enablerepo=fedora install systemd passwd yum fedora-release vim-minimal
 381 # systemd-nspawn -bD /srv/mycontainer</programlisting>
 382
 383                 <para>This installs a minimal Fedora distribution into
 384                 the directory <filename noindex='true'>/srv/mycontainer/</filename> and
 385                 then boots an OS in a namespace container in
 386                 it.</para>
 387         </refsect1>
 388
 389         <refsect1>
 390                 <title>Example 2</title>
 391
 392                 <programlisting># debootstrap --arch=amd64 unstable ~/debian-tree/
 393 # systemd-nspawn -D ~/debian-tree/</programlisting>
 394
 395                 <para>This installs a minimal Debian unstable
 396                 distribution into the directory
 397                 <filename>~/debian-tree/</filename> and then spawns a
 398                 shell in a namespace container in it.</para>
 399         </refsect1>
 400
 401         <refsect1>
 402                 <title>Example 3</title>
 403
 404                 <programlisting># pacstrap -c -d ~/arch-tree/ base
 405 # systemd-nspawn -bD ~/arch-tree/</programlisting>
 406
 407                 <para>This installs a mimimal Arch Linux distribution into
 408                 the directory <filename>~/arch-tree/</filename> and then
 409                 boots an OS in a namespace container in it.</para>
 410         </refsect1>
 411
 412         <refsect1 id='example-nsenter'>
 413                 <title>Example 4</title>
 414
 415                 <para>To enter the container, PID of one of the
 416                 processes sharing the new namespaces must be used.
 417                 <command>systemd-nspawn</command> prints the PID
 418                 (as viewed from the outside) of the launched process,
 419                 and it can be used to enter the container.</para>
 420
 421                 <programlisting># nsenter -m -u -i -n -p -t $PID</programlisting>
 422
 423                 <para><citerefentry><refentrytitle>nsenter</refentrytitle><manvolnum>1</manvolnum></citerefentry>
 424                 is part of
 425                 <ulink url="https://github.com/karelzak/util-linux">util-linux</ulink>.
 426                 Kernel support for entering namespaces was added in
 427                 Linux 3.8.</para>
 428         </refsect1>
 429
 430         <refsect1>
 431                 <title>Exit status</title>
 432
 433                 <para>The exit code of the program executed in the
 434                 container is returned.</para>
 435         </refsect1>
 436
 437         <refsect1>
 438                 <title>See Also</title>
 439                 <para>
 440                         <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
 441                         <citerefentry><refentrytitle>chroot</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
 442                         <citerefentry><refentrytitle>unshare</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
 443                         <citerefentry><refentrytitle>yum</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
 444                         <citerefentry><refentrytitle>debootstrap</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
 445                         <citerefentry><refentrytitle>pacman</refentrytitle><manvolnum>8</manvolnum></citerefentry>,
 446                         <citerefentry><refentrytitle>systemd.slice</refentrytitle><manvolnum>5</manvolnum></citerefentry>
 447                 </para>
 448         </refsect1>
 449
 450 </refentry>