1SYSTEMD-UDEVD.SERVICE(8) systemd-udevd.service SYSTEMD-UDEVD.SERVICE(8)
2
6 systemd-udevd.service, systemd-udevd-control.socket, systemd-udevd-
7 kernel.socket, systemd-udevd - Device event managing daemon
8
10 systemd-udevd.service
11 systemd-udevd-control.socket
13 systemd-udevd-kernel.socket
15 /usr/lib/systemd/systemd-udevd [--daemon] [--debug] [--children-max=]
17 [--exec-delay=] [--event-timeout=]
18 [--resolve-names=early|late|never]
19 [--version] [--help]
20
22 systemd-udevd listens to kernel uevents. For every event, systemd-udevd
23 executes matching instructions specified in udev rules. See udev(7).
24 The behavior of the daemon can be configured using udev.conf(5), its
26 command line options, environment variables, and on the kernel command
27 line, or changed dynamically with udevadm control.
28
30 -d, --daemon
31 Detach and run in the background.
32 -D, --debug
34 Print debug messages to standard error.
35 -c, --children-max=
37 Limit the number of events executed in parallel.
38 -e, --exec-delay=
40 Delay the execution of each RUN{program} parameter by the given
41 number of seconds. This option might be useful when debugging
42 system crashes during coldplug caused by loading non-working kernel
43 modules.
44 -t, --event-timeout=
46 Set the number of seconds to wait for events to finish. After this
47 time, the event will be terminated. The default is 180 seconds.
48 -s, --timeout-signal=
50 Set the signal which systemd-udevd will send to forked off
51 processes after reaching event timeout. The setting can be
52 overridden at boot time with the kernel command line option
53 udev.timeout_signal=. Setting to SIGABRT may be helpful in order to
54 debug worker timeouts. Defaults to SIGKILL. Note that setting the
55 option on the command line overrides the setting from the
56 configuration file.
57 -N, --resolve-names=
59 Specify when systemd-udevd should resolve names of users and
60 groups. When set to early (the default), names will be resolved
61 when the rules are parsed. When set to late, names will be resolved
62 for every event. When set to never, names will never be resolved
63 and all devices will be owned by root.
64 -h, --help
66 Print a short help text and exit.
67 --version
69 Print a short version string and exit.
70
72 Parameters prefixed with "rd." will be read when systemd-udevd is used
73 in an initrd, those without will be processed both in the initrd and on
74 the host.
75 udev.log_level=, rd.udev.log_level=
77 Set the log level.
78 udev.children_max=, rd.udev.children_max=
80 Limit the number of events executed in parallel.
81 udev.exec_delay=, rd.udev.exec_delay=
83 Delay the execution of each RUN{program} parameter by the given
84 number of seconds. This option might be useful when debugging
85 system crashes during coldplug caused by loading non-working kernel
86 modules.
87 udev.event_timeout=, rd.udev.event_timeout=
89 Wait for events to finish up to the given number of seconds. This
90 option might be useful if events are terminated due to kernel
91 drivers taking too long to initialize.
92 udev.timeout_signal=, rd.udev.timeout_signal=
94 Specifies a signal that systemd-udevd will send to workers on
95 timeout. Note that kernel command line option overrides both the
96 setting in the configuration file and the one on the program
97 command line.
98 udev.blockdev_read_only, rd.udev.blockdev_read_only
100 If specified, mark all physical block devices read-only as they
101 appear. Synthetic block devices (such as loopback block devices or
102 device mapper devices) are left as they are. This is useful to
103 guarantee that the contents of physical block devices remains
104 unmodified during runtime, for example to implement fully stateless
105 systems, for testing or for recovery situations where corrupted
106 file systems shall not be corrupted further through accidental
107 modification.
108 A block device may be marked writable again by issuing the blockdev
110 --setrw command, see blockdev(8) for details.
111 net.ifnames=
113 Network interfaces are renamed to give them predictable names when
114 possible. It is enabled by default; specifying 0 disables it.
115 net.naming-scheme=
117 Network interfaces are renamed to give them predictable names when
118 possible (unless net.ifnames=0 is specified, see above). With this
119 kernel command line option it is possible to pick a specific
120 version of this algorithm and override the default chosen at
121 compilation time. Expects one of the naming scheme identifiers
122 listed in systemd.net-naming-scheme(7), or "latest" to select the
123 latest scheme known (to this particular version of
124 systemd-udevd.service).
125 Note that selecting a specific scheme is not sufficient to fully
127 stabilize interface naming: the naming is generally derived from
128 driver attributes exposed by the kernel. As the kernel is updated,
129 previously missing attributes systemd-udevd.service is checking
130 might appear, which affects older name derivation algorithms, too.
131
133 udev.conf(5), udev(7), udevadm(8)
134systemd 248 SYSTEMD-UDEVD.SERVICE(8)