]> git.ipfire.org Git - thirdparty/systemd.git/blob - README
build-sys: ship fsck mock in tarball
[thirdparty/systemd.git] / README
1 systemd System and Service Manager
2
3 DETAILS:
4 http://0pointer.de/blog/projects/systemd.html
5
6 WEB SITE:
7 http://www.freedesktop.org/wiki/Software/systemd
8
9 GIT:
10 git://anongit.freedesktop.org/systemd/systemd
11 ssh://git.freedesktop.org/git/systemd/systemd
12
13 GITWEB:
14 http://cgit.freedesktop.org/systemd/systemd
15
16 MAILING LIST:
17 http://lists.freedesktop.org/mailman/listinfo/systemd-devel
18 http://lists.freedesktop.org/mailman/listinfo/systemd-commits
19
20 IRC:
21 #systemd on irc.freenode.org
22
23 BUG REPORTS:
24 https://bugs.freedesktop.org/enter_bug.cgi?product=systemd
25
26 AUTHOR:
27 Lennart Poettering
28 Kay Sievers
29 ...and many others
30
31 LICENSE:
32 LGPLv2.1+ for all code
33 - except src/shared/MurmurHash2.c which is Public Domain
34 - except src/shared/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+
37
38 REQUIREMENTS:
39 Linux kernel >= 3.7
40 Linux kernel >= 3.8 for Smack support
41
42 Kernel Config Options:
43 CONFIG_DEVTMPFS
44 CONFIG_CGROUPS (it is OK to disable all controllers)
45 CONFIG_INOTIFY_USER
46 CONFIG_SIGNALFD
47 CONFIG_TIMERFD
48 CONFIG_EPOLL
49 CONFIG_NET
50 CONFIG_SYSFS
51 CONFIG_PROC_FS
52 CONFIG_FHANDLE (libudev, mount and bind mount handling)
53
54 udev will fail to work with the legacy sysfs layout:
55 CONFIG_SYSFS_DEPRECATED=n
56
57 Legacy hotplug slows down the system and confuses udev:
58 CONFIG_UEVENT_HELPER_PATH=""
59
60 Userspace firmware loading is not supported and should
61 be disabled in the kernel:
62 CONFIG_FW_LOADER_USER_HELPER=n
63
64 Some udev rules and virtualization detection relies on it:
65 CONFIG_DMIID
66
67 Support for some SCSI devices serial number retrieval, to
68 create additional symlinks in /dev/disk/ and /dev/tape:
69 CONFIG_BLK_DEV_BSG
70
71 Required for PrivateNetwork and PrivateDevices in service units:
72 CONFIG_NET_NS
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.
76
77 Optional but strongly recommended:
78 CONFIG_IPV6
79 CONFIG_AUTOFS4_FS
80 CONFIG_TMPFS_XATTR
81 CONFIG_{TMPFS,EXT4,XFS,BTRFS_FS,...}_POSIX_ACL
82 CONFIG_SECCOMP
83 CONFIG_CHECKPOINT_RESTORE (for the kcmp() syscall)
84
85 Required for CPUShares in resource control unit settings
86 CONFIG_CGROUP_SCHED
87 CONFIG_FAIR_GROUP_SCHED
88
89 Required for CPUQuota in resource control unit settings
90 CONFIG_CFS_BANDWIDTH
91
92 For systemd-bootchart, several proc debug interfaces are required:
93 CONFIG_SCHEDSTATS
94 CONFIG_SCHED_DEBUG
95
96 For UEFI systems:
97 CONFIG_EFIVAR_FS
98 CONFIG_EFI_PARTITION
99
100 Note that kernel auditing is broken when used with systemd's
101 container code. When using systemd in conjunction with
102 containers, please make sure to either turn off auditing at
103 runtime using the kernel command line option "audit=0", or
104 turn it off at kernel compile time using:
105 CONFIG_AUDIT=n
106 If systemd is compiled with libseccomp support on
107 architectures which do not use socketcall() and where seccomp
108 is supported (this effectively means x86-64 and ARM, but
109 excludes 32-bit x86!), then nspawn will now install a
110 work-around seccomp filter that makes containers boot even
111 with audit being enabled. This works correctly only on kernels
112 3.14 and newer though. TL;DR: turn audit off, still.
113
114 glibc >= 2.16
115 libcap
116 libmount >= 2.20 (from util-linux)
117 libseccomp >= 1.0.0 (optional)
118 libblkid >= 2.24 (from util-linux) (optional)
119 libkmod >= 15 (optional)
120 PAM >= 1.1.2 (optional)
121 libcryptsetup (optional)
122 libaudit (optional)
123 libacl (optional)
124 libselinux (optional)
125 liblzma (optional)
126 liblz4 >= 119 (optional)
127 libgcrypt (optional)
128 libqrencode (optional)
129 libmicrohttpd (optional)
130 libpython (optional)
131 libidn (optional)
132 gobject-introspection > 1.40.0 (optional)
133 elfutils >= 158 (optional)
134 make, gcc, and similar tools
135
136 During runtime, you need the following additional
137 dependencies:
138
139 util-linux >= v2.26 required
140 dbus >= 1.4.0 (strictly speaking optional, but recommended)
141 dracut (optional)
142 PolicyKit (optional)
143
144 When building from git, you need the following additional
145 dependencies:
146
147 pkg-config
148 docbook-xsl
149 xsltproc
150 automake
151 autoconf
152 libtool
153 intltool
154 gperf
155 gtkdocize (optional)
156 python (optional)
157 python-lxml (optional, but required to build the indices)
158 sphinx (optional)
159
160 When systemd-hostnamed is used, it is strongly recommended to
161 install nss-myhostname to ensure that, in a world of
162 dynamically changing hostnames, the hostname stays resolvable
163 under all circumstances. In fact, systemd-hostnamed will warn
164 if nss-myhostname is not installed.
165
166 To build HTML documentation for python-systemd using sphinx,
167 please first install systemd (using 'make install'), and then
168 invoke sphinx-build with 'make sphinx-<target>', with <target>
169 being 'html' or 'latexpdf'. If using DESTDIR for installation,
170 pass the same DESTDIR to 'make sphinx-html' invocation.
171
172 USERS AND GROUPS:
173 Default udev rules use the following standard system group
174 names, which need to be resolvable by getgrnam() at any time,
175 even in the very early boot stages, where no other databases
176 and network are available:
177
178 audio, cdrom, dialout, disk, input, kmem, lp, tape, tty, video
179
180 During runtime, the journal daemon requires the
181 "systemd-journal" system group to exist. New journal files will
182 be readable by this group (but not writable), which may be used
183 to grant specific users read access. In addition, system
184 groups "wheel" and "adm" will be given read-only access to
185 journal files using systemd-tmpfiles.service.
186
187 The journal gateway daemon requires the
188 "systemd-journal-gateway" system user and group to
189 exist. During execution this network facing service will drop
190 privileges and assume this uid/gid for security reasons.
191
192 Similarly, the NTP daemon requires the "systemd-timesync" system
193 user and group to exist.
194
195 Similarly, the network management daemon requires the
196 "systemd-network" system user and group to exist.
197
198 Similarly, the name resolution daemon requires the
199 "systemd-resolve" system user and group to exist.
200
201 Similarly, the kdbus dbus1 proxy daemon requires the
202 "systemd-bus-proxy" system user and group to exist.
203
204 NSS:
205 systemd ships with three NSS modules:
206
207 nss-myhostname resolves the local hostname to locally
208 configured IP addresses, as well as "localhost" to
209 127.0.0.1/::1.
210
211 nss-resolve enables DNS resolution via the systemd-resolved
212 DNS/LLMNR caching stub resolver "systemd-resolved".
213
214 nss-mymachines enables resolution of all local containers
215 registered with machined to their respective IP addresses.
216
217 To make use of these NSS modules, please add them to the
218 "hosts: " line in /etc/nsswitch.conf. The "resolve" module
219 should replace the glibc "dns" module in this file.
220
221 The three modules should be used in the following order:
222
223 hosts: files mymachines resolve myhostname
224
225 WARNINGS:
226 systemd will warn you during boot if /etc/mtab is not a
227 symlink to /proc/mounts. Please ensure that /etc/mtab is a
228 proper symlink.
229
230 systemd will warn you during boot if /usr is on a different
231 file system than /. While in systemd itself very little will
232 break if /usr is on a separate partition, many of its
233 dependencies very likely will break sooner or later in one
234 form or another. For example, udev rules tend to refer to
235 binaries in /usr, binaries that link to libraries in /usr or
236 binaries that refer to data files in /usr. Since these
237 breakages are not always directly visible, systemd will warn
238 about this, since this kind of file system setup is not really
239 supported anymore by the basic set of Linux OS components.
240
241 systemd requires that the /run mount point exists. systemd also
242 requires that /var/run is a a symlink to /run.
243
244 For more information on this issue consult
245 http://freedesktop.org/wiki/Software/systemd/separate-usr-is-broken
246
247 To run systemd under valgrind, compile with VALGRIND defined
248 (e.g. ./configure CPPFLAGS='... -DVALGRIND=1'). Otherwise,
249 false positives will be triggered by code which violates
250 some rules but is actually safe.
251
252 ENGINEERING AND CONSULTING SERVICES:
253 ENDOCODE <https://endocode.com/> offers professional
254 engineering and consulting services for systemd. Please
255 contact Chris Kühl <chris@endocode.com> for more information.