1 systemd System and Service Manager
4 http://0pointer.de/blog/projects/systemd.html
7 https://www.freedesktop.org/wiki/Software/systemd
10 git@github.com:systemd/systemd.git
11 https://github.com/systemd/systemd.git
14 https://github.com/systemd/systemd
17 https://lists.freedesktop.org/mailman/listinfo/systemd-devel
20 #systemd on irc.freenode.org
23 https://github.com/systemd/systemd/issues
31 LGPLv2.1+ for all code
32 - except src/basic/MurmurHash2.c which is Public Domain
33 - except src/basic/siphash24.c which is CC0 Public Domain
34 - except src/journal/lookup3.c which is Public Domain
35 - except src/udev/* which is (currently still) GPLv2, GPLv2+
39 Linux kernel >= 4.2 for unified cgroup hierarchy support
41 Kernel Config Options:
43 CONFIG_CGROUPS (it is OK to disable all controllers)
51 CONFIG_FHANDLE (libudev, mount and bind mount handling)
53 Kernel crypto/hash API
54 CONFIG_CRYPTO_USER_API_HASH
58 udev will fail to work with the legacy sysfs layout:
59 CONFIG_SYSFS_DEPRECATED=n
61 Legacy hotplug slows down the system and confuses udev:
62 CONFIG_UEVENT_HELPER_PATH=""
64 Userspace firmware loading is not supported and should
65 be disabled in the kernel:
66 CONFIG_FW_LOADER_USER_HELPER=n
68 Some udev rules and virtualization detection relies on it:
71 Support for some SCSI devices serial number retrieval, to
72 create additional symlinks in /dev/disk/ and /dev/tape:
75 Required for PrivateNetwork= and PrivateDevices= in service units:
77 CONFIG_DEVPTS_MULTIPLE_INSTANCES
78 Note that systemd-localed.service and other systemd units use
79 PrivateNetwork and PrivateDevices so this is effectively required.
81 Required for PrivateUsers= in service units:
84 Optional but strongly recommended:
88 CONFIG_{TMPFS,EXT4,XFS,BTRFS_FS,...}_POSIX_ACL
90 CONFIG_SECCOMP_FILTER (required for seccomp support)
91 CONFIG_CHECKPOINT_RESTORE (for the kcmp() syscall)
93 Required for CPUShares= in resource control unit settings
95 CONFIG_FAIR_GROUP_SCHED
97 Required for CPUQuota= in resource control unit settings
104 We recommend to turn off Real-Time group scheduling in the
105 kernel when using systemd. RT group scheduling effectively
106 makes RT scheduling unavailable for most userspace, since it
107 requires explicit assignment of RT budgets to each unit whose
108 processes making use of RT. As there's no sensible way to
109 assign these budgets automatically this cannot really be
110 fixed, and it's best to disable group scheduling hence.
111 CONFIG_RT_GROUP_SCHED=n
113 Note that kernel auditing is broken when used with systemd's
114 container code. When using systemd in conjunction with
115 containers, please make sure to either turn off auditing at
116 runtime using the kernel command line option "audit=0", or
117 turn it off at kernel compile time using:
119 If systemd is compiled with libseccomp support on
120 architectures which do not use socketcall() and where seccomp
121 is supported (this effectively means x86-64 and ARM, but
122 excludes 32-bit x86!), then nspawn will now install a
123 work-around seccomp filter that makes containers boot even
124 with audit being enabled. This works correctly only on kernels
125 3.14 and newer though. TL;DR: turn audit off, still.
129 libmount >= 2.27.1 (from util-linux)
130 (util-linux < 2.29 *must* be built with --enable-libmount-force-mountinfo,
131 and later versions without --enable-libmount-support-mtab.)
132 libseccomp >= 2.3.1 (optional)
133 libblkid >= 2.24 (from util-linux) (optional)
134 libkmod >= 15 (optional)
135 PAM >= 1.1.2 (optional)
136 libcryptsetup (optional)
139 libselinux (optional)
141 liblz4 >= 119 (optional)
143 libqrencode (optional)
144 libmicrohttpd (optional)
146 libidn2 or libidn (optional)
147 elfutils >= 158 (optional)
148 make, gcc, and similar tools
150 During runtime, you need the following additional
153 util-linux >= v2.27.1 required
154 dbus >= 1.4.0 (strictly speaking optional, but recommended)
155 NOTE: If using dbus < 1.9.18, you should override the default
156 policy directory (--with-dbuspolicydir=/etc/dbus-1/system.d).
160 Two build systems are supported: meson + ninja-build and autools + make.
162 The following tools are needed with both systems:
166 docbook-xsl (optional, required for documentation)
167 xsltproc (optional, required for documentation)
168 python-lxml (optional, required to build the indices)
170 When building with meson, python and ninja-build are required.
172 To build in directory build/:
173 meson build/ && ninja -C build
175 Any configuration options can be specfied as -Darg=value... arguments
176 to meson. After the build directory is initially configured, meson will
177 refuse to run again, and options must be changed with:
178 mesonconf -Darg=value...
179 mesonconf without any arguments will print out available options and
180 their current values.
186 DESTDIR=... ninja install
188 When building with autotools, the following tools are needed:
196 The build system is initialized with ./autogen.sh and the usual
200 A tar ball can be created with:
201 git archive --format=tar --prefix=systemd-222/ v222 | xz > systemd-222.tar.xz
203 When systemd-hostnamed is used, it is strongly recommended to
204 install nss-myhostname to ensure that, in a world of
205 dynamically changing hostnames, the hostname stays resolvable
206 under all circumstances. In fact, systemd-hostnamed will warn
207 if nss-myhostname is not installed.
209 Additional packages are necessary to run some tests:
210 - busybox (used by test/TEST-13-NSPAWN-SMOKE)
211 - nc (used by test/TEST-12-ISSUE-3171)
213 - python3-evdev (used by hwdb parsing tests)
214 - strace (used by test/test-functions)
215 - capsh (optional, used by test-execute)
218 Default udev rules use the following standard system group
219 names, which need to be resolvable by getgrnam() at any time,
220 even in the very early boot stages, where no other databases
221 and network are available:
223 audio, cdrom, dialout, disk, input, kmem, lp, tape, tty, video
225 During runtime, the journal daemon requires the
226 "systemd-journal" system group to exist. New journal files will
227 be readable by this group (but not writable), which may be used
228 to grant specific users read access. In addition, system
229 groups "wheel" and "adm" will be given read-only access to
230 journal files using systemd-tmpfiles.service.
232 The journal gateway daemon requires the
233 "systemd-journal-gateway" system user and group to
234 exist. During execution this network facing service will drop
235 privileges and assume this uid/gid for security reasons.
237 Similarly, the NTP daemon requires the "systemd-timesync" system
238 user and group to exist.
240 Similarly, the network management daemon requires the
241 "systemd-network" system user and group to exist.
243 Similarly, the name resolution daemon requires the
244 "systemd-resolve" system user and group to exist.
246 Similarly, the coredump support requires the
247 "systemd-coredump" system user and group to exist.
250 systemd ships with four glibc NSS modules:
252 nss-myhostname resolves the local hostname to locally
253 configured IP addresses, as well as "localhost" to
256 nss-resolve enables DNS resolution via the systemd-resolved
257 DNS/LLMNR caching stub resolver "systemd-resolved".
259 nss-mymachines enables resolution of all local containers registered
260 with machined to their respective IP addresses. It also maps UID/GIDs
261 ranges used by containers to useful names.
263 nss-systemd enables resolution of all dynamically allocated service
264 users. (See the DynamicUser= setting in unit files.)
266 To make use of these NSS modules, please add them to the "hosts:",
267 "passwd:" and "group:" lines in /etc/nsswitch.conf. The "resolve"
268 module should replace the glibc "dns" module in this file (and don't
269 worry, it chain-loads the "dns" module if it can't talk to resolved).
271 The four modules should be used in the following order:
273 passwd: compat mymachines systemd
274 group: compat mymachines systemd
275 hosts: files mymachines resolve myhostname
278 When calling "systemctl enable/disable/is-enabled" on a unit which is a
279 SysV init.d script, it calls /usr/lib/systemd/systemd-sysv-install;
280 this needs to translate the action into the distribution specific
281 mechanism such as chkconfig or update-rc.d. Packagers need to provide
282 this script if you need this functionality (you don't if you disabled
285 Please see src/systemctl/systemd-sysv-install.SKELETON for how this
286 needs to look like, and provide an implementation at the marked places.
289 systemd will warn you during boot if /usr is on a different
290 file system than /. While in systemd itself very little will
291 break if /usr is on a separate partition, many of its
292 dependencies very likely will break sooner or later in one
293 form or another. For example, udev rules tend to refer to
294 binaries in /usr, binaries that link to libraries in /usr or
295 binaries that refer to data files in /usr. Since these
296 breakages are not always directly visible, systemd will warn
297 about this, since this kind of file system setup is not really
298 supported anymore by the basic set of Linux OS components.
300 systemd requires that the /run mount point exists. systemd also
301 requires that /var/run is a symlink to /run.
303 For more information on this issue consult
304 https://www.freedesktop.org/wiki/Software/systemd/separate-usr-is-broken
306 To run systemd under valgrind, compile with VALGRIND defined
307 (e.g. ./configure CPPFLAGS='... -DVALGRIND=1'). Otherwise,
308 false positives will be triggered by code which violates
309 some rules but is actually safe.
311 ENGINEERING AND CONSULTING SERVICES:
312 Kinvolk (https://kinvolk.io) offers professional engineering
313 and consulting services for systemd. Please contact Chris Kühl
314 <chris@kinvolk.io> for more information.