1 <?xml version='
1.0'
?> <!--*- Mode: nxml; nxml-child-indent: 2; indent-tabs-mode: nil -*-->
2 <!DOCTYPE refentry PUBLIC
"-//OASIS//DTD DocBook XML V4.2//EN"
3 "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
6 SPDX-License-Identifier: LGPL-2.1+
8 This file is part of systemd.
10 Copyright 2012 Lennart Poettering
12 systemd is free software; you can redistribute it and/or modify it
13 under the terms of the GNU Lesser General Public License as published by
14 the Free Software Foundation; either version 2.1 of the License, or
15 (at your option) any later version.
17 systemd is distributed in the hope that it will be useful, but
18 WITHOUT ANY WARRANTY; without even the implied warranty of
19 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 Lesser General Public License for more details.
22 You should have received a copy of the GNU Lesser General Public License
23 along with systemd; If not, see <http://www.gnu.org/licenses/>.
26 <refentry id=
"systemd-cgtop"
27 xmlns:
xi=
"http://www.w3.org/2001/XInclude">
30 <title>systemd-cgtop
</title>
31 <productname>systemd
</productname>
35 <contrib>Developer
</contrib>
36 <firstname>Lennart
</firstname>
37 <surname>Poettering
</surname>
38 <email>lennart@poettering.net
</email>
44 <refentrytitle>systemd-cgtop
</refentrytitle>
45 <manvolnum>1</manvolnum>
49 <refname>systemd-cgtop
</refname>
50 <refpurpose>Show top control groups by their resource usage
</refpurpose>
55 <command>systemd-cgtop
</command>
56 <arg choice=
"opt" rep=
"repeat">OPTIONS
</arg>
57 <arg choice=
"opt">GROUP
</arg>
62 <title>Description
</title>
64 <para><command>systemd-cgtop
</command> shows the top control
65 groups of the local Linux control group hierarchy, ordered by
66 their CPU, memory, or disk I/O load. The display is refreshed in
67 regular intervals (by default every
1s), similar in style to
68 <citerefentry project='man-pages'
><refentrytitle>top
</refentrytitle><manvolnum>1</manvolnum></citerefentry>.
69 If a control group path is specified, shows only the services of
70 the specified control group.
</para>
72 <para>If
<command>systemd-cgtop
</command> is not connected to a
73 tty, no column headers are printed and the default is to only run
74 one iteration. The
<varname>--iterations=
</varname> argument, if
75 given, is honored. This mode is suitable for scripting.
</para>
77 <para>Resource usage is only accounted for control groups in the
78 relevant hierarchy, i.e. CPU usage is only accounted for control
79 groups in the
<literal>cpuacct
</literal> hierarchy, memory usage
80 only for those in
<literal>memory
</literal> and disk I/O usage for
81 those in
<literal>blkio
</literal>. If resource monitoring for
82 these resources is required, it is recommended to add the
83 <varname>CPUAccounting=
1</varname>,
84 <varname>MemoryAccounting=
1</varname> and
85 <varname>BlockIOAccounting=
1</varname> settings in the unit files
87 <citerefentry><refentrytitle>systemd.resource-control
</refentrytitle><manvolnum>5</manvolnum></citerefentry>
90 <para>The CPU load value can be between
0 and
100 times the number of
91 processors the system has. For example, if the system has
8 processors,
92 the CPU load value is going to be between
0% and
800%. The number of
93 processors can be found in
<literal>/proc/cpuinfo
</literal>.
</para>
95 <para>To emphasize this: unless
96 <literal>CPUAccounting=
1</literal>,
97 <literal>MemoryAccounting=
1</literal> and
98 <literal>BlockIOAccounting=
1</literal> are enabled for the
99 services in question, no resource accounting will be available for
100 system services and the data shown by
101 <command>systemd-cgtop
</command> will be incomplete.
</para>
105 <title>Options
</title>
107 <para>The following options are understood:
</para>
111 <term><option>-p
</option></term>
112 <term><option>--order=path
</option></term>
114 <listitem><para>Order by control group
115 path name.
</para></listitem>
119 <term><option>-t
</option></term>
120 <term><option>--order=tasks
</option></term>
122 <listitem><para>Order by number of tasks/processes in the control group.
</para></listitem>
126 <term><option>-c
</option></term>
127 <term><option>--order=cpu
</option></term>
129 <listitem><para>Order by CPU load.
</para></listitem>
133 <term><option>-m
</option></term>
134 <term><option>--order=memory
</option></term>
136 <listitem><para>Order by memory usage.
</para></listitem>
140 <term><option>-i
</option></term>
141 <term><option>--order=io
</option></term>
143 <listitem><para>Order by disk I/O load.
</para></listitem>
147 <term><option>-b
</option></term>
148 <term><option>--batch
</option></term>
150 <listitem><para>Run in
"batch" mode: do not accept input and
151 run until the iteration limit set with
152 <option>--iterations=
</option> is exhausted or until killed.
153 This mode could be useful for sending output from
154 <command>systemd-cgtop
</command> to other programs or to a
155 file.
</para></listitem>
159 <term><option>-r
</option></term>
160 <term><option>--raw
</option></term>
162 <listitem><para>Format byte counts (as in memory usage and I/O metrics)
163 with raw numeric values rather than human-readable
164 numbers.
</para></listitem>
168 <term><option>--cpu=percentage
</option></term>
169 <term><option>--cpu=time
</option></term>
171 <listitem><para>Controls whether the CPU usage is shown as
172 percentage or time. By default, the CPU usage is shown as
173 percentage. This setting may also be toggled at runtime by
174 pressing the
<keycap>%
</keycap> key.
</para></listitem>
178 <term><option>-P
</option></term>
180 <listitem><para>Count only userspace processes instead of all
181 tasks. By default, all tasks are counted: each kernel thread
182 and each userspace thread individually. With this setting,
183 kernel threads are excluded from the counting and each
184 userspace process only counts as one, regardless how many
185 threads it consists of. This setting may also be toggled at
186 runtime by pressing the
<keycap>P
</keycap> key. This option
187 may not be combined with
188 <option>-k
</option>.
</para></listitem>
192 <term><option>-k
</option></term>
194 <listitem><para>Count only userspace processes and kernel
195 threads instead of all tasks. By default, all tasks are
196 counted: each kernel thread and each userspace thread
197 individually. With this setting, kernel threads are included in
198 the counting and each userspace process only counts as on one,
199 regardless how many threads it consists of. This setting may
200 also be toggled at runtime by pressing the
<keycap>k
</keycap>
201 key. This option may not be combined with
202 <option>-P
</option>.
</para></listitem>
206 <term><option>--recursive=
</option></term>
208 <listitem><para>Controls whether the number of processes shown
209 for a control group shall include all processes that are
210 contained in any of the child control groups as well. Takes a
211 boolean argument, which defaults to
<literal>yes
</literal>. If
212 enabled, the processes in child control groups are included, if
213 disabled, only the processes in the control group itself are
214 counted. This setting may also be toggled at runtime by
215 pressing the
<keycap>r
</keycap> key. Note that this setting
216 only applies to process counting, i.e. when the
217 <option>-P
</option> or
<option>-k
</option> options are
218 used. It has not effect if all tasks are counted, in which
219 case the counting is always recursive.
</para></listitem>
223 <term><option>-n
</option></term>
224 <term><option>--iterations=
</option></term>
226 <listitem><para>Perform only this many iterations. A value of
227 0 indicates that the program should run
228 indefinitely.
</para></listitem>
232 <term><option>-
1</option></term>
234 <listitem><para>A shortcut for
<option>--iterations=
1</option>.
</para></listitem>
238 <term><option>-d
</option></term>
239 <term><option>--delay=
</option></term>
241 <listitem><para>Specify refresh delay in seconds (or if one of
242 <literal>ms
</literal>,
<literal>us
</literal>,
243 <literal>min
</literal> is specified as unit in this time
244 unit). This setting may also be increased and decreased at
245 runtime by pressing the
<keycap>+
</keycap> and
246 <keycap>-
</keycap> keys.
</para></listitem>
250 <term><option>--depth=
</option></term>
252 <listitem><para>Maximum control group tree traversal depth.
253 Specifies how deep
<command>systemd-cgtop
</command> shall
254 traverse the control group hierarchies. If
0 is specified,
255 only the root group is monitored. For
1, only the first level
256 of control groups is monitored, and so on. Defaults to
261 <term><option>-M
<replaceable>MACHINE
</replaceable></option></term>
262 <term><option>--machine=
<replaceable>MACHINE
</replaceable></option></term>
264 <listitem><para>Limit control groups shown to the part
265 corresponding to the container
266 <replaceable>MACHINE
</replaceable>.
267 This option may not be used when a control group path is specified.
</para></listitem>
270 <xi:include href=
"standard-options.xml" xpointer=
"help" />
271 <xi:include href=
"standard-options.xml" xpointer=
"version" />
279 <para><command>systemd-cgtop
</command> is an interactive tool and
280 may be controlled via user input using the following keys:
</para>
284 <term><keycap>h
</keycap></term>
286 <listitem><para>Shows a short help text.
</para></listitem>
290 <term><keycap function=
"space"/></term>
292 <listitem><para>Immediately refresh output.
</para></listitem>
296 <term><keycap>q
</keycap></term>
298 <listitem><para>Terminate the program.
</para></listitem>
302 <term><keycap>p
</keycap></term>
303 <term><keycap>t
</keycap></term>
304 <term><keycap>c
</keycap></term>
305 <term><keycap>m
</keycap></term>
306 <term><keycap>i
</keycap></term>
308 <listitem><para>Sort the control groups by path, number of
309 tasks, CPU load, memory usage, or I/O load, respectively. This
310 setting may also be controlled using the
311 <option>--order=
</option> command line
312 switch.
</para></listitem>
316 <term><keycap>%
</keycap></term>
318 <listitem><para>Toggle between showing CPU time as time or
319 percentage. This setting may also be controlled using the
320 <option>--cpu=
</option> command line switch.
</para></listitem>
324 <term><keycap>+
</keycap></term>
325 <term><keycap>-
</keycap></term>
327 <listitem><para>Increase or decrease refresh delay,
328 respectively. This setting may also be controlled using the
329 <option>--delay=
</option> command line
330 switch.
</para></listitem>
334 <term><keycap>P
</keycap></term>
336 <listitem><para>Toggle between counting all tasks, or only
337 userspace processes. This setting may also be controlled using
338 the
<option>-P
</option> command line switch (see
339 above).
</para></listitem>
343 <term><keycap>k
</keycap></term>
345 <listitem><para>Toggle between counting all tasks, or only
346 userspace processes and kernel threads. This setting may also
347 be controlled using the
<option>-k
</option> command line
348 switch (see above).
</para></listitem>
352 <term><keycap>r
</keycap></term>
354 <listitem><para>Toggle between recursively including or
355 excluding processes in child control groups in control group
356 process counts. This setting may also be controlled using the
357 <option>--recursive=
</option> command line switch. This key is
358 not available if all tasks are counted, it is only available
359 if processes are counted, as enabled with the
360 <keycap>P
</keycap> or
<keycap>k
</keycap>
361 keys.
</para></listitem>
368 <title>Exit status
</title>
370 <para>On success,
0 is returned, a non-zero failure code
375 <title>See Also
</title>
377 <citerefentry><refentrytitle>systemd
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
378 <citerefentry><refentrytitle>systemctl
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
379 <citerefentry><refentrytitle>systemd-cgls
</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
380 <citerefentry><refentrytitle>systemd.resource-control
</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
381 <citerefentry project='man-pages'
><refentrytitle>top
</refentrytitle><manvolnum>1</manvolnum></citerefentry>