1 systemd System and Service Manager
4 http://0pointer.de/blog/projects/systemd.html
7 http://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 http://lists.freedesktop.org/mailman/listinfo/systemd-devel
18 http://lists.freedesktop.org/mailman/listinfo/systemd-commits
21 #systemd on irc.freenode.org
24 https://github.com/systemd/systemd/issues
32 LGPLv2.1+ for all code
33 - except src/basic/MurmurHash2.c which is Public Domain
34 - except src/basic/siphash24.c which is CC0 Public Domain
35 - except src/journal/lookup3.c which is Public Domain
36 - except src/udev/* which is (currently still) GPLv2, GPLv2+
40 Linux kernel >= 4.2 for unified cgroup hierarchy support
42 Kernel Config Options:
44 CONFIG_CGROUPS (it is OK to disable all controllers)
52 CONFIG_FHANDLE (libudev, mount and bind mount handling)
54 udev will fail to work with the legacy sysfs layout:
55 CONFIG_SYSFS_DEPRECATED=n
57 Legacy hotplug slows down the system and confuses udev:
58 CONFIG_UEVENT_HELPER_PATH=""
60 Userspace firmware loading is not supported and should
61 be disabled in the kernel:
62 CONFIG_FW_LOADER_USER_HELPER=n
64 Some udev rules and virtualization detection relies on it:
67 Support for some SCSI devices serial number retrieval, to
68 create additional symlinks in /dev/disk/ and /dev/tape:
71 Required for PrivateNetwork and PrivateDevices in service units:
73 CONFIG_DEVPTS_MULTIPLE_INSTANCES
74 Note that systemd-localed.service and other systemd units use
75 PrivateNetwork and PrivateDevices so this is effectively required.
77 Optional but strongly recommended:
81 CONFIG_{TMPFS,EXT4,XFS,BTRFS_FS,...}_POSIX_ACL
83 CONFIG_CHECKPOINT_RESTORE (for the kcmp() syscall)
85 Required for CPUShares= in resource control unit settings
87 CONFIG_FAIR_GROUP_SCHED
89 Required for CPUQuota= in resource control unit settings
92 For systemd-bootchart, several proc debug interfaces are required:
100 We recommend to turn off Real-Time group scheduling in the
101 kernel when using systemd. RT group scheduling effectively
102 makes RT scheduling unavailable for most userspace, since it
103 requires explicit assignment of RT budgets to each unit whose
104 processes making use of RT. As there's no sensible way to
105 assign these budgets automatically this cannot really be
106 fixed, and it's best to disable group scheduling hence.
107 CONFIG_RT_GROUP_SCHED=n
109 Note that kernel auditing is broken when used with systemd's
110 container code. When using systemd in conjunction with
111 containers, please make sure to either turn off auditing at
112 runtime using the kernel command line option "audit=0", or
113 turn it off at kernel compile time using:
115 If systemd is compiled with libseccomp support on
116 architectures which do not use socketcall() and where seccomp
117 is supported (this effectively means x86-64 and ARM, but
118 excludes 32-bit x86!), then nspawn will now install a
119 work-around seccomp filter that makes containers boot even
120 with audit being enabled. This works correctly only on kernels
121 3.14 and newer though. TL;DR: turn audit off, still.
125 libmount >= 2.27.1 (from util-linux)
126 libseccomp >= 1.0.0 (optional)
127 libblkid >= 2.24 (from util-linux) (optional)
128 libkmod >= 15 (optional)
129 PAM >= 1.1.2 (optional)
130 libcryptsetup (optional)
133 libselinux (optional)
135 liblz4 >= 119 (optional)
137 libqrencode (optional)
138 libmicrohttpd (optional)
141 elfutils >= 158 (optional)
142 make, gcc, and similar tools
144 During runtime, you need the following additional
147 util-linux >= v2.27.1 required
148 dbus >= 1.4.0 (strictly speaking optional, but recommended)
152 When building from git, the following tools are needed:
163 python-lxml (optional, but required to build the indices)
165 The build system is initialized with ./autogen.sh. A tar ball
167 git archive --format=tar --prefix=systemd-222/ v222 | xz > systemd-222.tar.xz
169 When systemd-hostnamed is used, it is strongly recommended to
170 install nss-myhostname to ensure that, in a world of
171 dynamically changing hostnames, the hostname stays resolvable
172 under all circumstances. In fact, systemd-hostnamed will warn
173 if nss-myhostname is not installed.
176 Default udev rules use the following standard system group
177 names, which need to be resolvable by getgrnam() at any time,
178 even in the very early boot stages, where no other databases
179 and network are available:
181 audio, cdrom, dialout, disk, input, kmem, lp, tape, tty, video
183 During runtime, the journal daemon requires the
184 "systemd-journal" system group to exist. New journal files will
185 be readable by this group (but not writable), which may be used
186 to grant specific users read access. In addition, system
187 groups "wheel" and "adm" will be given read-only access to
188 journal files using systemd-tmpfiles.service.
190 The journal gateway daemon requires the
191 "systemd-journal-gateway" system user and group to
192 exist. During execution this network facing service will drop
193 privileges and assume this uid/gid for security reasons.
195 Similarly, the NTP daemon requires the "systemd-timesync" system
196 user and group to exist.
198 Similarly, the network management daemon requires the
199 "systemd-network" system user and group to exist.
201 Similarly, the name resolution daemon requires the
202 "systemd-resolve" system user and group to exist.
204 Similarly, the kdbus dbus1 proxy daemon requires the
205 "systemd-bus-proxy" system user and group to exist.
208 systemd ships with three NSS modules:
210 nss-myhostname resolves the local hostname to locally
211 configured IP addresses, as well as "localhost" to
214 nss-resolve enables DNS resolution via the systemd-resolved
215 DNS/LLMNR caching stub resolver "systemd-resolved".
217 nss-mymachines enables resolution of all local containers
218 registered with machined to their respective IP addresses.
220 To make use of these NSS modules, please add them to the
221 "hosts: " line in /etc/nsswitch.conf. The "resolve" module
222 should replace the glibc "dns" module in this file.
224 The three modules should be used in the following order:
226 hosts: files mymachines resolve myhostname
229 When calling "systemctl enable/disable/is-enabled" on a unit which is a
230 SysV init.d script, it calls /usr/lib/systemd/systemd-sysv-install;
231 this needs to translate the action into the distribution specific
232 mechanism such as chkconfig or update-rc.d. Packagers need to provide
233 this script if you need this functionality (you don't if you disabled
236 Please see src/systemctl/systemd-sysv-install.SKELETON for how this
237 needs to look like, and provide an implementation at the marked places.
240 systemd will freeze execution during boot if /etc/mtab exists
241 but is not a symlink to /proc/mounts. Please ensure that
242 /etc/mtab is a proper symlink.
244 systemd will warn you during boot if /usr is on a different
245 file system than /. While in systemd itself very little will
246 break if /usr is on a separate partition, many of its
247 dependencies very likely will break sooner or later in one
248 form or another. For example, udev rules tend to refer to
249 binaries in /usr, binaries that link to libraries in /usr or
250 binaries that refer to data files in /usr. Since these
251 breakages are not always directly visible, systemd will warn
252 about this, since this kind of file system setup is not really
253 supported anymore by the basic set of Linux OS components.
255 systemd requires that the /run mount point exists. systemd also
256 requires that /var/run is a symlink to /run.
258 For more information on this issue consult
259 http://freedesktop.org/wiki/Software/systemd/separate-usr-is-broken
261 To run systemd under valgrind, compile with VALGRIND defined
262 (e.g. ./configure CPPFLAGS='... -DVALGRIND=1'). Otherwise,
263 false positives will be triggered by code which violates
264 some rules but is actually safe.
266 Currently, systemd-timesyncd defaults to use the Google NTP
267 servers if not specified otherwise at configure time. You
268 really should not ship an OS or device with this default
269 setting. See DISTRO_PORTING for details.
271 ENGINEERING AND CONSULTING SERVICES:
272 Kinvolk (https://kinvolk.io) offers professional engineering
273 and consulting services for systemd. Please contact Chris Kühl
274 <chris@kinvolk.io> for more information.