2 <!DOCTYPE refentry PUBLIC
"-//OASIS//DTD DocBook XML V4.5//EN"
3 "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
4 <!-- SPDX-License-Identifier: LGPL-2.1+ -->
6 <refentry id=
"systemd.resource-control">
8 <title>systemd.resource-control
</title>
9 <productname>systemd
</productname>
13 <refentrytitle>systemd.resource-control
</refentrytitle>
14 <manvolnum>5</manvolnum>
18 <refname>systemd.resource-control
</refname>
19 <refpurpose>Resource control unit settings
</refpurpose>
24 <filename><replaceable>slice
</replaceable>.slice
</filename>,
25 <filename><replaceable>scope
</replaceable>.scope
</filename>,
26 <filename><replaceable>service
</replaceable>.service
</filename>,
27 <filename><replaceable>socket
</replaceable>.socket
</filename>,
28 <filename><replaceable>mount
</replaceable>.mount
</filename>,
29 <filename><replaceable>swap
</replaceable>.swap
</filename>
34 <title>Description
</title>
36 <para>Unit configuration files for services, slices, scopes, sockets, mount points, and swap devices share a subset
37 of configuration options for resource control of spawned processes. Internally, this relies on the Linux Control
38 Groups (cgroups) kernel concept for organizing processes in a hierarchical tree of named groups for the purpose of
39 resource management.
</para>
41 <para>This man page lists the configuration options shared by
42 those six unit types. See
43 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
44 for the common options of all unit configuration files, and
45 <citerefentry><refentrytitle>systemd.slice
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
46 <citerefentry><refentrytitle>systemd.scope
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
47 <citerefentry><refentrytitle>systemd.service
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
48 <citerefentry><refentrytitle>systemd.socket
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
49 <citerefentry><refentrytitle>systemd.mount
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
51 <citerefentry><refentrytitle>systemd.swap
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
52 for more information on the specific unit configuration files. The
53 resource control configuration options are configured in the
54 [Slice], [Scope], [Service], [Socket], [Mount], or [Swap]
55 sections, depending on the unit type.
</para>
57 <para>In addition, options which control resources available to programs
58 <emphasis>executed
</emphasis> by systemd are listed in
59 <citerefentry><refentrytitle>systemd.exec
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
60 Those options complement options listed here.
</para>
63 url=
"https://www.freedesktop.org/wiki/Software/systemd/ControlGroupInterface/">New
64 Control Group Interfaces
</ulink> for an introduction on how to make
65 use of resource control APIs from programs.
</para>
69 <title>Implicit Dependencies
</title>
71 <para>The following dependencies are implicitly added:
</para>
74 <listitem><para>Units with the
<varname>Slice=
</varname> setting set automatically acquire
75 <varname>Requires=
</varname> and
<varname>After=
</varname> dependencies on the specified
76 slice unit.
</para></listitem>
80 <!-- We don't have any default dependency here. -->
83 <title>Unified and Legacy Control Group Hierarchies
</title>
85 <para>The unified control group hierarchy is the new version of kernel control group interface, see
<ulink
86 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>. Depending on the resource type,
87 there are differences in resource control capabilities. Also, because of interface changes, some resource types
88 have separate set of options on the unified hierarchy.
</para>
94 <term><option>CPU
</option></term>
96 <para><varname>CPUWeight=
</varname> and
<varname>StartupCPUWeight=
</varname> replace
97 <varname>CPUShares=
</varname> and
<varname>StartupCPUShares=
</varname>, respectively.
</para>
99 <para>The
<literal>cpuacct
</literal> controller does not exist separately on the unified hierarchy.
</para>
104 <term><option>Memory
</option></term>
106 <para><varname>MemoryMax=
</varname> replaces
<varname>MemoryLimit=
</varname>.
<varname>MemoryLow=
</varname>
107 and
<varname>MemoryHigh=
</varname> are effective only on unified hierarchy.
</para>
112 <term><option>IO
</option></term>
114 <para><varname>IO
</varname> prefixed settings are a superset of and replace
<varname>BlockIO
</varname>
115 prefixed ones. On unified hierarchy, IO resource control also applies to buffered writes.
</para>
122 <para>To ease the transition, there is best-effort translation between the two versions of settings. For each
123 controller, if any of the settings for the unified hierarchy are present, all settings for the legacy hierarchy are
124 ignored. If the resulting settings are for the other type of hierarchy, the configurations are translated before
127 <para>Legacy control group hierarchy (see
<ulink
128 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/cgroups.txt">cgroups.txt
</ulink>), also called cgroup-v1,
129 doesn't allow safe delegation of controllers to unprivileged processes. If the system uses the legacy control group
130 hierarchy, resource control is disabled for systemd user instance, see
131 <citerefentry><refentrytitle>systemd
</refentrytitle><manvolnum>1</manvolnum></citerefentry>.
136 <title>Options
</title>
138 <para>Units of the types listed above can have settings
139 for resource control configuration:
</para>
141 <variablelist class='unit-directives'
>
144 <term><varname>CPUAccounting=
</varname></term>
147 <para>Turn on CPU usage accounting for this unit. Takes a
148 boolean argument. Note that turning on CPU accounting for
149 one unit will also implicitly turn it on for all units
150 contained in the same slice and for all its parent slices
151 and the units contained therein. The system default for this
152 setting may be controlled with
153 <varname>DefaultCPUAccounting=
</varname> in
154 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
159 <term><varname>CPUWeight=
<replaceable>weight
</replaceable></varname></term>
160 <term><varname>StartupCPUWeight=
<replaceable>weight
</replaceable></varname></term>
163 <para>Assign the specified CPU time weight to the processes executed, if the unified control group hierarchy
164 is used on the system. These options take an integer value and control the
<literal>cpu.weight
</literal>
165 control group attribute. The allowed range is
1 to
10000. Defaults to
100. For details about this control
166 group attribute, see
<ulink
167 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink> and
<ulink
168 url=
"https://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt">sched-design-CFS.txt
</ulink>.
169 The available CPU time is split up among all units within one slice relative to their CPU time weight.
</para>
171 <para>While
<varname>StartupCPUWeight=
</varname> only applies to the startup phase of the system,
172 <varname>CPUWeight=
</varname> applies to normal runtime of the system, and if the former is not set also to
173 the startup phase. Using
<varname>StartupCPUWeight=
</varname> allows prioritizing specific services at
174 boot-up differently than during normal runtime.
</para>
176 <para>These settings replace
<varname>CPUShares=
</varname> and
<varname>StartupCPUShares=
</varname>.
</para>
181 <term><varname>CPUQuota=
</varname></term>
184 <para>Assign the specified CPU time quota to the processes executed. Takes a percentage value, suffixed with
185 "%". The percentage specifies how much CPU time the unit shall get at maximum, relative to the total CPU time
186 available on one CPU. Use values
> 100% for allotting CPU time on more than one CPU. This controls the
187 <literal>cpu.max
</literal> attribute on the unified control group hierarchy and
188 <literal>cpu.cfs_quota_us
</literal> on legacy. For details about these control group attributes, see
<ulink
189 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink> and
<ulink
190 url=
"https://www.kernel.org/doc/Documentation/scheduler/sched-bwc.txt">sched-bwc.txt
</ulink>.
</para>
192 <para>Example:
<varname>CPUQuota=
20%
</varname> ensures that the executed processes will never get more than
193 20% CPU time on one CPU.
</para>
199 <term><varname>CPUQuotaPeriodSec=
</varname></term>
202 <para>Assign the duration over which the CPU time quota specified by
<varname>CPUQuota=
</varname> is measured.
203 Takes a time duration value in seconds, with an optional suffix such as
"ms" for milliseconds (or
"s" for seconds.)
204 The default setting is
100ms. The period is clamped to the range supported by the kernel, which is [
1ms,
1000ms].
205 Additionally, the period is adjusted up so that the quota interval is also at least
1ms.
206 Setting
<varname>CPUQuotaPeriodSec=
</varname> to an empty value resets it to the default.
</para>
208 <para>This controls the second field of
<literal>cpu.max
</literal> attribute on the unified control group hierarchy
209 and
<literal>cpu.cfs_period_us
</literal> on legacy. For details about these control group attributes, see
210 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink> and
211 <ulink url=
"https://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt">sched-design-CFS.txt
</ulink>.
</para>
213 <para>Example:
<varname>CPUQuotaPeriodSec=
10ms
</varname> to request that the CPU quota is measured in periods of
10ms.
</para>
218 <term><varname>MemoryAccounting=
</varname></term>
221 <para>Turn on process and kernel memory accounting for this
222 unit. Takes a boolean argument. Note that turning on memory
223 accounting for one unit will also implicitly turn it on for
224 all units contained in the same slice and for all its parent
225 slices and the units contained therein. The system default
226 for this setting may be controlled with
227 <varname>DefaultMemoryAccounting=
</varname> in
228 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
233 <term><varname>MemoryMin=
<replaceable>bytes
</replaceable></varname></term>
236 <para>Specify the memory usage protection of the executed processes in this unit. If the memory usages of
237 this unit and all its ancestors are below their minimum boundaries, this unit's memory won't be reclaimed.
</para>
239 <para>Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is
240 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base
1024), respectively. Alternatively, a
241 percentage value may be specified, which is taken relative to the installed physical memory on the
242 system. This controls the
<literal>memory.min
</literal> control group attribute. For details about this
243 control group attribute, see
<ulink
244 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
246 <para>This setting is supported only if the unified control group hierarchy is used and disables
247 <varname>MemoryLimit=
</varname>.
</para>
252 <term><varname>MemoryLow=
<replaceable>bytes
</replaceable></varname></term>
255 <para>Specify the best-effort memory usage protection of the executed processes in this unit. If the memory
256 usages of this unit and all its ancestors are below their low boundaries, this unit's memory won't be
257 reclaimed as long as memory can be reclaimed from unprotected units.
</para>
259 <para>Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is
260 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base
1024), respectively. Alternatively, a
261 percentage value may be specified, which is taken relative to the installed physical memory on the
262 system. This controls the
<literal>memory.low
</literal> control group attribute. For details about this
263 control group attribute, see
<ulink
264 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
266 <para>This setting is supported only if the unified control group hierarchy is used and disables
267 <varname>MemoryLimit=
</varname>.
</para>
272 <term><varname>MemoryHigh=
<replaceable>bytes
</replaceable></varname></term>
275 <para>Specify the high limit on memory usage of the executed processes in this unit. Memory usage may go
276 above the limit if unavoidable, but the processes are heavily slowed down and memory is taken away
277 aggressively in such cases. This is the main mechanism to control memory usage of a unit.
</para>
279 <para>Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is
280 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base
1024), respectively. Alternatively, a
281 percentage value may be specified, which is taken relative to the installed physical memory on the
282 system. If assigned the
283 special value
<literal>infinity
</literal>, no memory limit is applied. This controls the
284 <literal>memory.high
</literal> control group attribute. For details about this control group attribute, see
285 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
287 <para>This setting is supported only if the unified control group hierarchy is used and disables
288 <varname>MemoryLimit=
</varname>.
</para>
293 <term><varname>MemoryMax=
<replaceable>bytes
</replaceable></varname></term>
296 <para>Specify the absolute limit on memory usage of the executed processes in this unit. If memory usage
297 cannot be contained under the limit, out-of-memory killer is invoked inside the unit. It is recommended to
298 use
<varname>MemoryHigh=
</varname> as the main control mechanism and use
<varname>MemoryMax=
</varname> as the
299 last line of defense.
</para>
301 <para>Takes a memory size in bytes. If the value is suffixed with K, M, G or T, the specified memory size is
302 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base
1024), respectively. Alternatively, a
303 percentage value may be specified, which is taken relative to the installed physical memory on the system. If
304 assigned the special value
<literal>infinity
</literal>, no memory limit is applied. This controls the
305 <literal>memory.max
</literal> control group attribute. For details about this control group attribute, see
306 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
308 <para>This setting replaces
<varname>MemoryLimit=
</varname>.
</para>
313 <term><varname>MemorySwapMax=
<replaceable>bytes
</replaceable></varname></term>
316 <para>Specify the absolute limit on swap usage of the executed processes in this unit.
</para>
318 <para>Takes a swap size in bytes. If the value is suffixed with K, M, G or T, the specified swap size is
319 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base
1024), respectively. If assigned the
320 special value
<literal>infinity
</literal>, no swap limit is applied. This controls the
321 <literal>memory.swap.max
</literal> control group attribute. For details about this control group attribute,
322 see
<ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
324 <para>This setting is supported only if the unified control group hierarchy is used and disables
325 <varname>MemoryLimit=
</varname>.
</para>
330 <term><varname>TasksAccounting=
</varname></term>
333 <para>Turn on task accounting for this unit. Takes a
334 boolean argument. If enabled, the system manager will keep
335 track of the number of tasks in the unit. The number of
336 tasks accounted this way includes both kernel threads and
337 userspace processes, with each thread counting
338 individually. Note that turning on tasks accounting for one
339 unit will also implicitly turn it on for all units contained
340 in the same slice and for all its parent slices and the
341 units contained therein. The system default for this setting
342 may be controlled with
343 <varname>DefaultTasksAccounting=
</varname> in
344 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
349 <term><varname>TasksMax=
<replaceable>N
</replaceable></varname></term>
352 <para>Specify the maximum number of tasks that may be created in the unit. This ensures that the number of
353 tasks accounted for the unit (see above) stays below a specific limit. This either takes an absolute number
354 of tasks or a percentage value that is taken relative to the configured maximum number of tasks on the
355 system. If assigned the special value
<literal>infinity
</literal>, no tasks limit is applied. This controls
356 the
<literal>pids.max
</literal> control group attribute. For details about this control group attribute, see
357 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/pids.txt">pids.txt
</ulink>.
</para>
360 system default for this setting may be controlled with
361 <varname>DefaultTasksMax=
</varname> in
362 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
367 <term><varname>IOAccounting=
</varname></term>
370 <para>Turn on Block I/O accounting for this unit, if the unified control group hierarchy is used on the
371 system. Takes a boolean argument. Note that turning on block I/O accounting for one unit will also implicitly
372 turn it on for all units contained in the same slice and all for its parent slices and the units contained
373 therein. The system default for this setting may be controlled with
<varname>DefaultIOAccounting=
</varname>
375 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
377 <para>This setting replaces
<varname>BlockIOAccounting=
</varname> and disables settings prefixed with
378 <varname>BlockIO
</varname> or
<varname>StartupBlockIO
</varname>.
</para>
383 <term><varname>IOWeight=
<replaceable>weight
</replaceable></varname></term>
384 <term><varname>StartupIOWeight=
<replaceable>weight
</replaceable></varname></term>
387 <para>Set the default overall block I/O weight for the executed processes, if the unified control group
388 hierarchy is used on the system. Takes a single weight value (between
1 and
10000) to set the default block
389 I/O weight. This controls the
<literal>io.weight
</literal> control group attribute, which defaults to
390 100. For details about this control group attribute, see
<ulink
391 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>. The available I/O
392 bandwidth is split up among all units within one slice relative to their block I/O weight.
</para>
394 <para>While
<varname>StartupIOWeight=
</varname> only applies
395 to the startup phase of the system,
396 <varname>IOWeight=
</varname> applies to the later runtime of
397 the system, and if the former is not set also to the startup
398 phase. This allows prioritizing specific services at boot-up
399 differently than during runtime.
</para>
401 <para>These settings replace
<varname>BlockIOWeight=
</varname> and
<varname>StartupBlockIOWeight=
</varname>
402 and disable settings prefixed with
<varname>BlockIO
</varname> or
<varname>StartupBlockIO
</varname>.
</para>
407 <term><varname>IODeviceWeight=
<replaceable>device
</replaceable> <replaceable>weight
</replaceable></varname></term>
410 <para>Set the per-device overall block I/O weight for the executed processes, if the unified control group
411 hierarchy is used on the system. Takes a space-separated pair of a file path and a weight value to specify
412 the device specific weight value, between
1 and
10000. (Example:
<literal>/dev/sda
1000</literal>). The file
413 path may be specified as path to a block device node or as any other file, in which case the backing block
414 device of the file system of the file is determined. This controls the
<literal>io.weight
</literal> control
415 group attribute, which defaults to
100. Use this option multiple times to set weights for multiple devices.
416 For details about this control group attribute, see
<ulink
417 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
419 <para>This setting replaces
<varname>BlockIODeviceWeight=
</varname> and disables settings prefixed with
420 <varname>BlockIO
</varname> or
<varname>StartupBlockIO
</varname>.
</para>
425 <term><varname>IOReadBandwidthMax=
<replaceable>device
</replaceable> <replaceable>bytes
</replaceable></varname></term>
426 <term><varname>IOWriteBandwidthMax=
<replaceable>device
</replaceable> <replaceable>bytes
</replaceable></varname></term>
429 <para>Set the per-device overall block I/O bandwidth maximum limit for the executed processes, if the unified
430 control group hierarchy is used on the system. This limit is not work-conserving and the executed processes
431 are not allowed to use more even if the device has idle capacity. Takes a space-separated pair of a file
432 path and a bandwidth value (in bytes per second) to specify the device specific bandwidth. The file path may
433 be a path to a block device node, or as any other file in which case the backing block device of the file
434 system of the file is used. If the bandwidth is suffixed with K, M, G, or T, the specified bandwidth is
435 parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes, respectively, to the base of
1000. (Example:
436 "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 5M"). This controls the
<literal>io.max
</literal> control
437 group attributes. Use this option multiple times to set bandwidth limits for multiple devices. For details
438 about this control group attribute, see
<ulink
439 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
442 <para>These settings replace
<varname>BlockIOReadBandwidth=
</varname> and
443 <varname>BlockIOWriteBandwidth=
</varname> and disable settings prefixed with
<varname>BlockIO
</varname> or
444 <varname>StartupBlockIO
</varname>.
</para>
449 <term><varname>IOReadIOPSMax=
<replaceable>device
</replaceable> <replaceable>IOPS
</replaceable></varname></term>
450 <term><varname>IOWriteIOPSMax=
<replaceable>device
</replaceable> <replaceable>IOPS
</replaceable></varname></term>
453 <para>Set the per-device overall block I/O IOs-Per-Second maximum limit for the executed processes, if the
454 unified control group hierarchy is used on the system. This limit is not work-conserving and the executed
455 processes are not allowed to use more even if the device has idle capacity. Takes a space-separated pair of
456 a file path and an IOPS value to specify the device specific IOPS. The file path may be a path to a block
457 device node, or as any other file in which case the backing block device of the file system of the file is
458 used. If the IOPS is suffixed with K, M, G, or T, the specified IOPS is parsed as KiloIOPS, MegaIOPS,
459 GigaIOPS, or TeraIOPS, respectively, to the base of
1000. (Example:
460 "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 1K"). This controls the
<literal>io.max
</literal> control
461 group attributes. Use this option multiple times to set IOPS limits for multiple devices. For details about
462 this control group attribute, see
<ulink
463 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
466 <para>These settings are supported only if the unified control group hierarchy is used and disable settings
467 prefixed with
<varname>BlockIO
</varname> or
<varname>StartupBlockIO
</varname>.
</para>
472 <term><varname>IODeviceLatencyTargetSec=
<replaceable>device
</replaceable> <replaceable>target
</replaceable></varname></term>
475 <para>Set the per-device average target I/O latency for the executed processes, if the unified control group
476 hierarchy is used on the system. Takes a file path and a timespan separated by a space to specify
477 the device specific latency target. (Example:
"/dev/sda 25ms"). The file path may be specified
478 as path to a block device node or as any other file, in which case the backing block device of the file
479 system of the file is determined. This controls the
<literal>io.latency
</literal> control group
480 attribute. Use this option multiple times to set latency target for multiple devices. For details about this
481 control group attribute, see
<ulink
482 url=
"https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt
</ulink>.
</para>
484 <para>Implies
<literal>IOAccounting=yes
</literal>.
</para>
486 <para>These settings are supported only if the unified control group hierarchy is used.
</para>
491 <term><varname>IPAccounting=
</varname></term>
494 <para>Takes a boolean argument. If true, turns on IPv4 and IPv6 network traffic accounting for packets sent
495 or received by the unit. When this option is turned on, all IPv4 and IPv6 sockets created by any process of
496 the unit are accounted for.
</para>
498 <para>When this option is used in socket units, it applies to all IPv4 and IPv6 sockets
499 associated with it (including both listening and connection sockets where this applies). Note that for
500 socket-activated services, this configuration setting and the accounting data of the service unit and the
501 socket unit are kept separate, and displayed separately. No propagation of the setting and the collected
502 statistics is done, in either direction. Moreover, any traffic sent or received on any of the socket unit's
503 sockets is accounted to the socket unit — and never to the service unit it might have activated, even if the
504 socket is used by it.
</para>
506 <para>The system default for this setting may be controlled with
<varname>DefaultIPAccounting=
</varname> in
507 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
512 <term><varname>IPAddressAllow=
<replaceable>ADDRESS[/PREFIXLENGTH]…
</replaceable></varname></term>
513 <term><varname>IPAddressDeny=
<replaceable>ADDRESS[/PREFIXLENGTH]…
</replaceable></varname></term>
516 <para>Turn on address range network traffic filtering for packets sent and received over AF_INET and AF_INET6
517 sockets. Both directives take a space separated list of IPv4 or IPv6 addresses, each optionally suffixed
518 with an address prefix length (separated by a
<literal>/
</literal> character). If the latter is omitted, the
519 address is considered a host address, i.e. the prefix covers the whole address (
32 for IPv4,
128 for IPv6).
522 <para>The access lists configured with this option are applied to all sockets created by processes of this
523 unit (or in the case of socket units, associated with it). The lists are implicitly combined with any lists
524 configured for any of the parent slice units this unit might be a member of. By default all access lists are
525 empty. When configured the lists are enforced as follows:
</para>
528 <listitem><para>Access will be granted in case its destination/source address matches any entry in the
529 <varname>IPAddressAllow=
</varname> setting.
</para></listitem>
531 <listitem><para>Otherwise, access will be denied in case its destination/source address matches any entry
532 in the
<varname>IPAddressDeny=
</varname> setting.
</para></listitem>
534 <listitem><para>Otherwise, access will be granted.
</para></listitem>
537 <para>In order to implement a whitelisting IP firewall, it is recommended to use a
538 <varname>IPAddressDeny=
</varname><constant>any
</constant> setting on an upper-level slice unit (such as the
539 root slice
<filename>-.slice
</filename> or the slice containing all system services
540 <filename>system.slice
</filename> – see
541 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>7</manvolnum></citerefentry> for
542 details on these slice units), plus individual per-service
<varname>IPAddressAllow=
</varname> lines
543 permitting network access to relevant services, and only them.
</para>
545 <para>Note that for socket-activated services, the IP access list configured on the socket unit applies to
546 all sockets associated with it directly, but not to any sockets created by the ultimately activated services
547 for it. Conversely, the IP access list configured for the service is not applied to any sockets passed into
548 the service via socket activation. Thus, it is usually a good idea, to replicate the IP access lists on both
549 the socket and the service unit, however it often makes sense to maintain one list more open and the other
550 one more restricted, depending on the usecase.
</para>
552 <para>If these settings are used multiple times in the same unit the specified lists are combined. If an
553 empty string is assigned to these settings the specific access list is reset and all previous settings undone.
</para>
555 <para>In place of explicit IPv4 or IPv6 address and prefix length specifications a small set of symbolic
556 names may be used. The following names are defined:
</para>
559 <title>Special address/network names
</title>
562 <colspec colname='name'
/>
563 <colspec colname='definition'
/>
564 <colspec colname='meaning'
/>
568 <entry>Symbolic Name
</entry>
569 <entry>Definition
</entry>
570 <entry>Meaning
</entry>
576 <entry><constant>any
</constant></entry>
577 <entry>0.0.0.0/
0 ::/
0</entry>
578 <entry>Any host
</entry>
582 <entry><constant>localhost
</constant></entry>
583 <entry>127.0.0.0/
8 ::
1/
128</entry>
584 <entry>All addresses on the local loopback
</entry>
588 <entry><constant>link-local
</constant></entry>
589 <entry>169.254.0.0/
16 fe80::/
64</entry>
590 <entry>All link-local IP addresses
</entry>
594 <entry><constant>multicast
</constant></entry>
595 <entry>224.0.0.0/
4 ff00::/
8</entry>
596 <entry>All IP multicasting addresses
</entry>
602 <para>Note that these settings might not be supported on some systems (for example if eBPF control group
603 support is not enabled in the underlying kernel or container manager). These settings will have no effect in
604 that case. If compatibility with such systems is desired it is hence recommended to not exclusively rely on
605 them for IP security.
</para>
610 <term><varname>DeviceAllow=
</varname></term>
613 <para>Control access to specific device nodes by the
614 executed processes. Takes two space-separated strings: a
615 device node specifier followed by a combination of
616 <constant>r
</constant>,
<constant>w
</constant>,
617 <constant>m
</constant> to control
618 <emphasis>r
</emphasis>eading,
<emphasis>w
</emphasis>riting,
619 or creation of the specific device node(s) by the unit
620 (
<emphasis>m
</emphasis>knod), respectively. This controls
621 the
<literal>devices.allow
</literal> and
622 <literal>devices.deny
</literal> control group
623 attributes. For details about these control group
624 attributes, see
<ulink
625 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/devices.txt">devices.txt
</ulink>.
</para>
627 <para>The device node specifier is either a path to a device
628 node in the file system, starting with
629 <filename>/dev/
</filename>, or a string starting with either
630 <literal>char-
</literal> or
<literal>block-
</literal>
631 followed by a device group name, as listed in
632 <filename>/proc/devices
</filename>. The latter is useful to
633 whitelist all current and future devices belonging to a
634 specific device group at once. The device group is matched
635 according to filename globbing rules, you may hence use the
636 <literal>*
</literal> and
<literal>?
</literal>
637 wildcards. Examples:
<filename>/dev/sda5
</filename> is a
638 path to a device node, referring to an ATA or SCSI block
639 device.
<literal>char-pts
</literal> and
640 <literal>char-alsa
</literal> are specifiers for all pseudo
641 TTYs and all ALSA sound devices,
642 respectively.
<literal>char-cpu/*
</literal> is a specifier
643 matching all CPU related device groups.
</para>
648 <term><varname>DevicePolicy=auto|closed|strict
</varname></term>
652 Control the policy for allowing device access:
656 <term><option>strict
</option></term>
658 <para>means to only allow types of access that are
659 explicitly specified.
</para>
664 <term><option>closed
</option></term>
666 <para>in addition, allows access to standard pseudo
668 <filename>/dev/null
</filename>,
669 <filename>/dev/zero
</filename>,
670 <filename>/dev/full
</filename>,
671 <filename>/dev/random
</filename>, and
672 <filename>/dev/urandom
</filename>.
678 <term><option>auto
</option></term>
681 in addition, allows access to all devices if no
682 explicit
<varname>DeviceAllow=
</varname> is present.
692 <term><varname>Slice=
</varname></term>
695 <para>The name of the slice unit to place the unit
696 in. Defaults to
<filename>system.slice
</filename> for all
697 non-instantiated units of all unit types (except for slice
698 units themselves see below). Instance units are by default
699 placed in a subslice of
<filename>system.slice
</filename>
700 that is named after the template name.
</para>
702 <para>This option may be used to arrange systemd units in a
703 hierarchy of slices each of which might have resource
704 settings applied.
</para>
706 <para>For units of type slice, the only accepted value for
707 this setting is the parent slice. Since the name of a slice
708 unit implies the parent slice, it is hence redundant to ever
709 set this parameter directly for slice units.
</para>
711 <para>Special care should be taken when relying on the default slice assignment in templated service units
712 that have
<varname>DefaultDependencies=no
</varname> set, see
713 <citerefentry><refentrytitle>systemd.service
</refentrytitle><manvolnum>5</manvolnum></citerefentry>, section
714 "Default Dependencies" for details.
</para>
720 <term><varname>Delegate=
</varname></term>
723 <para>Turns on delegation of further resource control partitioning to processes of the unit. Units where this
724 is enabled may create and manage their own private subhierarchy of control groups below the control group of
725 the unit itself. For unprivileged services (i.e. those using the
<varname>User=
</varname> setting) the unit's
726 control group will be made accessible to the relevant user. When enabled the service manager will refrain
727 from manipulating control groups or moving processes below the unit's control group, so that a clear concept
728 of ownership is established: the control group tree above the unit's control group (i.e. towards the root
729 control group) is owned and managed by the service manager of the host, while the control group tree below
730 the unit's control group is owned and managed by the unit itself. Takes either a boolean argument or a list
731 of control group controller names. If true, delegation is turned on, and all supported controllers are
732 enabled for the unit, making them available to the unit's processes for management. If false, delegation is
733 turned off entirely (and no additional controllers are enabled). If set to a list of controllers, delegation
734 is turned on, and the specified controllers are enabled for the unit. Note that additional controllers than
735 the ones specified might be made available as well, depending on configuration of the containing slice unit
736 or other units contained in it. Note that assigning the empty string will enable delegation, but reset the
737 list of controllers, all assignments prior to this will have no effect. Defaults to false.
</para>
739 <para>Note that controller delegation to less privileged code is only safe on the unified control group
740 hierarchy. Accordingly, access to the specified controllers will not be granted to unprivileged services on
741 the legacy hierarchy, even when requested.
</para>
743 <para>The following controller names may be specified:
<option>cpu
</option>,
<option>cpuacct
</option>,
744 <option>io
</option>,
<option>blkio
</option>,
<option>memory
</option>,
<option>devices
</option>,
745 <option>pids
</option>. Not all of these controllers are available on all kernels however, and some are
746 specific to the unified hierarchy while others are specific to the legacy hierarchy. Also note that the
747 kernel might support further controllers, which aren't covered here yet as delegation is either not supported
748 at all for them or not defined cleanly.
</para>
750 <para>For further details on the delegation model consult
<ulink
751 url=
"https://systemd.io/CGROUP_DELEGATION">Control Group APIs and Delegation
</ulink>.
</para>
756 <term><varname>DisableControllers=
</varname></term>
759 <para>Disables controllers from being enabled for a unit's children. If a controller listed is already in use
760 in its subtree, the controller will be removed from the subtree. This can be used to avoid child units being
761 able to implicitly or explicitly enable a controller. Defaults to not disabling any controllers.
</para>
763 <para>It may not be possible to successfully disable a controller if the unit or any child of the unit in
764 question delegates controllers to its children, as any delegated subtree of the cgroup hierarchy is unmanaged
767 <para>Multiple controllers may be specified, separated by spaces. You may also pass
768 <varname>DisableControllers=
</varname> multiple times, in which case each new instance adds another controller
769 to disable. Passing
<varname>DisableControllers=
</varname> by itself with no controller name present resets
770 the disabled controller list.
</para>
772 <para>Valid controllers are
<option>cpu
</option>,
<option>cpuacct
</option>,
<option>io
</option>,
773 <option>blkio
</option>,
<option>memory
</option>,
<option>devices
</option>, and
<option>pids
</option>.
</para>
780 <title>Deprecated Options
</title>
782 <para>The following options are deprecated. Use the indicated superseding options instead:
</para>
784 <variablelist class='unit-directives'
>
787 <term><varname>CPUShares=
<replaceable>weight
</replaceable></varname></term>
788 <term><varname>StartupCPUShares=
<replaceable>weight
</replaceable></varname></term>
791 <para>Assign the specified CPU time share weight to the processes executed. These options take an integer
792 value and control the
<literal>cpu.shares
</literal> control group attribute. The allowed range is
2 to
793 262144. Defaults to
1024. For details about this control group attribute, see
<ulink
794 url=
"https://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt">sched-design-CFS.txt
</ulink>.
795 The available CPU time is split up among all units within one slice relative to their CPU time share
798 <para>While
<varname>StartupCPUShares=
</varname> only applies to the startup phase of the system,
799 <varname>CPUShares=
</varname> applies to normal runtime of the system, and if the former is not set also to
800 the startup phase. Using
<varname>StartupCPUShares=
</varname> allows prioritizing specific services at
801 boot-up differently than during normal runtime.
</para>
803 <para>Implies
<literal>CPUAccounting=yes
</literal>.
</para>
805 <para>These settings are deprecated. Use
<varname>CPUWeight=
</varname> and
806 <varname>StartupCPUWeight=
</varname> instead.
</para>
811 <term><varname>MemoryLimit=
<replaceable>bytes
</replaceable></varname></term>
814 <para>Specify the limit on maximum memory usage of the executed processes. The limit specifies how much
815 process and kernel memory can be used by tasks in this unit. Takes a memory size in bytes. If the value is
816 suffixed with K, M, G or T, the specified memory size is parsed as Kilobytes, Megabytes, Gigabytes, or
817 Terabytes (with the base
1024), respectively. Alternatively, a percentage value may be specified, which is
818 taken relative to the installed physical memory on the system. If assigned the special value
819 <literal>infinity
</literal>, no memory limit is applied. This controls the
820 <literal>memory.limit_in_bytes
</literal> control group attribute. For details about this control group
821 attribute, see
<ulink
822 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/memory.txt">memory.txt
</ulink>.
</para>
824 <para>Implies
<literal>MemoryAccounting=yes
</literal>.
</para>
826 <para>This setting is deprecated. Use
<varname>MemoryMax=
</varname> instead.
</para>
831 <term><varname>BlockIOAccounting=
</varname></term>
834 <para>Turn on Block I/O accounting for this unit, if the legacy control group hierarchy is used on the
835 system. Takes a boolean argument. Note that turning on block I/O accounting for one unit will also implicitly
836 turn it on for all units contained in the same slice and all for its parent slices and the units contained
837 therein. The system default for this setting may be controlled with
838 <varname>DefaultBlockIOAccounting=
</varname> in
839 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>.
</para>
841 <para>This setting is deprecated. Use
<varname>IOAccounting=
</varname> instead.
</para>
846 <term><varname>BlockIOWeight=
<replaceable>weight
</replaceable></varname></term>
847 <term><varname>StartupBlockIOWeight=
<replaceable>weight
</replaceable></varname></term>
849 <listitem><para>Set the default overall block I/O weight for the executed processes, if the legacy control
850 group hierarchy is used on the system. Takes a single weight value (between
10 and
1000) to set the default
851 block I/O weight. This controls the
<literal>blkio.weight
</literal> control group attribute, which defaults to
852 500. For details about this control group attribute, see
<ulink
853 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/blkio-controller.txt">blkio-controller.txt
</ulink>.
854 The available I/O bandwidth is split up among all units within one slice relative to their block I/O
857 <para>While
<varname>StartupBlockIOWeight=
</varname> only
858 applies to the startup phase of the system,
859 <varname>BlockIOWeight=
</varname> applies to the later runtime
860 of the system, and if the former is not set also to the
861 startup phase. This allows prioritizing specific services at
862 boot-up differently than during runtime.
</para>
865 <literal>BlockIOAccounting=yes
</literal>.
</para>
867 <para>These settings are deprecated. Use
<varname>IOWeight=
</varname> and
<varname>StartupIOWeight=
</varname>
874 <term><varname>BlockIODeviceWeight=
<replaceable>device
</replaceable> <replaceable>weight
</replaceable></varname></term>
877 <para>Set the per-device overall block I/O weight for the executed processes, if the legacy control group
878 hierarchy is used on the system. Takes a space-separated pair of a file path and a weight value to specify
879 the device specific weight value, between
10 and
1000. (Example:
"/dev/sda 500"). The file path may be
880 specified as path to a block device node or as any other file, in which case the backing block device of the
881 file system of the file is determined. This controls the
<literal>blkio.weight_device
</literal> control group
882 attribute, which defaults to
1000. Use this option multiple times to set weights for multiple devices. For
883 details about this control group attribute, see
<ulink
884 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/blkio-controller.txt">blkio-controller.txt
</ulink>.
</para>
887 <literal>BlockIOAccounting=yes
</literal>.
</para>
889 <para>This setting is deprecated. Use
<varname>IODeviceWeight=
</varname> instead.
</para>
894 <term><varname>BlockIOReadBandwidth=
<replaceable>device
</replaceable> <replaceable>bytes
</replaceable></varname></term>
895 <term><varname>BlockIOWriteBandwidth=
<replaceable>device
</replaceable> <replaceable>bytes
</replaceable></varname></term>
898 <para>Set the per-device overall block I/O bandwidth limit for the executed processes, if the legacy control
899 group hierarchy is used on the system. Takes a space-separated pair of a file path and a bandwidth value (in
900 bytes per second) to specify the device specific bandwidth. The file path may be a path to a block device
901 node, or as any other file in which case the backing block device of the file system of the file is used. If
902 the bandwidth is suffixed with K, M, G, or T, the specified bandwidth is parsed as Kilobytes, Megabytes,
903 Gigabytes, or Terabytes, respectively, to the base of
1000. (Example:
904 "/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 5M"). This controls the
905 <literal>blkio.throttle.read_bps_device
</literal> and
<literal>blkio.throttle.write_bps_device
</literal>
906 control group attributes. Use this option multiple times to set bandwidth limits for multiple devices. For
907 details about these control group attributes, see
<ulink
908 url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/blkio-controller.txt">blkio-controller.txt
</ulink>.
912 <literal>BlockIOAccounting=yes
</literal>.
</para>
914 <para>These settings are deprecated. Use
<varname>IOReadBandwidthMax=
</varname> and
915 <varname>IOWriteBandwidthMax=
</varname> instead.
</para>
923 <title>See Also
</title>
925 <citerefentry><refentrytitle>systemd
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
926 <citerefentry><refentrytitle>systemd-system.conf
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
927 <citerefentry><refentrytitle>systemd.unit
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
928 <citerefentry><refentrytitle>systemd.service
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
929 <citerefentry><refentrytitle>systemd.slice
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
930 <citerefentry><refentrytitle>systemd.scope
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
931 <citerefentry><refentrytitle>systemd.socket
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
932 <citerefentry><refentrytitle>systemd.mount
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
933 <citerefentry><refentrytitle>systemd.swap
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
934 <citerefentry><refentrytitle>systemd.exec
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
935 <citerefentry><refentrytitle>systemd.directives
</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
936 <citerefentry><refentrytitle>systemd.special
</refentrytitle><manvolnum>7</manvolnum></citerefentry>,
937 The documentation for control groups and specific controllers in the Linux kernel:
938 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/cgroups.txt">cgroups.txt
</ulink>,
939 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/cpuacct.txt">cpuacct.txt
</ulink>,
940 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/memory.txt">memory.txt
</ulink>,
941 <ulink url=
"https://www.kernel.org/doc/Documentation/cgroup-v1/blkio-controller.txt">blkio-controller.txt
</ulink>.
942 <ulink url=
"https://www.kernel.org/doc/Documentation/scheduler/sched-bwc.txt">sched-bwc.txt
</ulink>.