userfaultfd.2: Add Linux container migration use-case to NOTES

[man-pages.git] / man7 / cgroups.7

.\" Copyright (C) 2015 Serge Hallyn <serge@hallyn.com>
.\" and Copyright (C) 2016 Michael Kerrisk <mtk.manpages@gmail.com>
.\"
.\" %%%LICENSE_START(VERBATIM)
.\" Permission is granted to make and distribute verbatim copies of this
.\" manual provided the copyright notice and this permission notice are
.\" preserved on all copies.
.\"
.\" Permission is granted to copy and distribute modified versions of this
.\" manual under the conditions for verbatim copying, provided that the
.\" entire resulting derived work is distributed under the terms of a
.\" permission notice identical to this one.
.\"
.\" Since the Linux kernel and libraries are constantly changing, this
.\" manual page may be incorrect or out-of-date.  The author(s) assume no
.\" responsibility for errors or omissions, or for damages resulting from
.\" the use of the information contained herein.  The author(s) may not
.\" have taken the same level of care in the production of this manual,
.\" which is licensed free of charge, as they might when working
.\" professionally.
.\"
.\" Formatted or processed versions of this manual, if unaccompanied by
.\" the source, must acknowledge the copyright and authors of this work.
.\" %%%LICENSE_END
.\"
.TH CGROUPS 7 2016-12-12 "Linux" "Linux Programmer's Manual"
.SH NAME
cgroups \- Linux control groups
.SH DESCRIPTION
Control cgroups, usually referred to as cgroups,
are a Linux kernel feature which allow processes to
be organized into hierarchical groups whose usage of
various types of resources can then be limited and monitored.
The kernel's cgroup interface is provided through
a pseudo-filesystem called cgroupfs.
Grouping is implemented in the core cgroup kernel code,
while resource tracking and limits are implemented in
a set of per-resource-type subsystems (memory, CPU, and so on).
.\"
.SS Terminology
A
.I cgroup
is a collection of processes that are bound to a set of
limits or parameters defined via the cgroup filesystem.

A
.I subsystem
is a kernel component that modifies the behavior of
the processes in a cgroup.
Various subsystems have been implemented, making it possible to do things
such as limiting the amount of CPU time and memory available to a cgroup,
accounting for the CPU time used by a cgroup,
and freezing and resuming execution of the processes in a cgroup.
Subsystems are sometimes also known as
.IR "resource controllers"
(or simply, controllers).

The cgroups for a controller are arranged in a
.IR hierarchy .
This hierarchy is defined by creating, removing, and
renaming subdirectories within the cgroup filesystem.
At each level of the hierarchy, attributes (e.g., limits) can be defined.
The limits, control, and accounting provided by cgroups generally have
effect throughout the subhierarchy underneath the cgroup where the
attributes are defined.
Thus, for example, the limits placed on
a cgroup at a higher level in the hierarchy cannot be exceeded
by descendant cgroups.
.\"
.SS Cgroups version 1 and version 2
The initial release of the cgroups implementation was in Linux 2.6.24.
Over time, various cgroup controllers have been added
to allow the management of various types of resources.
However, the development of these controllers was largely uncoordinated,
with the result that many inconsistencies arose between controllers
and management of the cgroup hierarchies became rather complex.
(A longer description of these problems can be found in
the kernel source file
.IR Documentation/cgroup\-v2.txt .)

Because of the problems with the initial cgroups implementation
(cgroups version 1),
starting in Linux 3.10, work began on a new,
orthogonal implementation to remedy these problems.
Initially marked experimental, and hidden behind the
.I "\-o\ __DEVEL__sane_behavior"
mount option, the new version (cgroups version 2)
was eventually made official with the release of Linux 4.5.
Differences between the two versions are described in the text below.

Although cgroups v2 is intended as a replacement for cgroups v1,
the older system continues to exist
(and for compatibility reasons is unlikely to be removed).
Currently, cgroups v2 implements only a subset of the controllers
available in cgroups v1.
The two systems are implemented so that both v1 controllers and
v2 controllers can be mounted on the same system.
Thus, for example, it is possible to use those controllers
that are supported under version 2,
while also using version 1 controllers
where version 2 does not yet support those controllers.
The only restriction here is that a controller can't be simultaneously
employed in both a cgroups v1 hierarchy and in the cgroups v2 hierarchy.
.\"
.SS Cgroups version 1
Under cgroups v1, each controller may be mounted against a separate
cgroup filesystem that provides its own hierarchical organization of the
processes on the system.
It is also possible comount multiple (or even all) cgroups v1 controllers
against the same cgroup filesystem, meaning that the comounted controllers
manage the same hierarchical organization of processes.

For each mounted hierarchy,
the directory tree mirrors the control group hierarchy.
Each control group is represented by a directory, with each of its child
control cgroups represented as a child directory.
For instance,
.IR /user/joe/1.session
represents control group
.IR 1.session ,
which is a child of cgroup
.IR joe ,
which is a child of
.IR /user .
Under each cgroup directory is a set of files which can be read or
written to, reflecting resource limits and a few general cgroup
properties.

In addition, in cgroups v1,
cgroups can be mounted with no bound controller, in which case
they serve only to track processes.
(See the discussion of release notification below.)
An example of this is the
.I name=systemd
cgroup which is used by
.BR systemd (1)
to track services and user sessions.
.\"
.SS Tasks (threads) versus processes
In cgroups v1, a distinction is drawn between
.I processes
and
.IR tasks .
In this view, a process can consist of multiple tasks
(more commonly called threads, from a user-space perspective,
and called such in the remainder of this man page).
In cgroups v1, it is possible to independently manipulate
the cgroup memberships of the threads in a process.
Because this ability caused certain problems,
.\" FIXME Add some text describing why this was a problem.
the ability to independently manipulate the cgroup memberships
of the threads in a process has been removed in cgroups v2.
Cgroups v2 allows manipulation of cgroup membership only for processes
(which has the effect of changing the cgroup membership of
all threads in the process).
.\"
.SS Mounting v1 controllers
The use of cgroups requires a kernel built with the
.BR CONFIG_CGROUP
option.
In addition, each of the v1 controllers has an associated
configuration option that must be set in order to employ that controller.

In order to use a v1 controller,
it must be mounted against a cgroup filesystem.
The usual place for such mounts is under a
.BR tmpfs (5)
filesystem mounted at
.IR /sys/fs/cgroup .
Thus, one might mount the
.I cpu
controller as follows:

.nf
.in +4n
mount \-t cgroup \-o cpu none /sys/fs/cgroup/cpu
.in
.fi

It is possible to comount multiple controllers against the same hierarchy.
For example, here the
.IR cpu
and
.IR cpuacct
controllers are comounted against a single hierarchy:

.nf
.in +4n
mount \-t cgroup \-o cpu,cpuacct none /sys/fs/cgroup/cpu,cpuacct
.in
.fi

Comounting controllers has the effect that a process is in the same cgroup for
all of the comounted controllers.
Separately mounting controllers allows a process to
be in cgroup
.I /foo1
for one controller while being in
.I /foo2/foo3
for another.

It is possible to comount all v1 controllers against the same hierarchy:

.nf
.in +4n
mount \-t cgroup \-o all cgroup /sys/fs/cgroup
.in
.fi

(One can achieve the same result by omitting
.IR "\-o all" ,
since it is the default if no controllers are explicitly specified.)

It is not possible to mount the same controller
against multiple cgroup hierarchies.
For example, it is not possible to mount both the
.I cpu
and
.I cpuacct
controllers against one hierarchy, and to mount the
.I cpu
controller alone against another hierarchy.
It is possible to create multiple mount points with exactly
the same set of comounted controllers.
However, in this case all that results is multiple mount points
providing a view of the same hierarchy.

Note that on many systems, the v1 controllers are automatically mounted under
.IR /sys/fs/cgroup ;
in particular,
.BR systemd (1)
automatically creates such mount points.
.\"
.SS Cgroups version 1 controllers
Each of the cgroups version 1 controllers is governed
by a kernel configuration option (listed below).
Additionally, the availability of the cgroups feature is governed by the
.BR CONFIG_CGROUPS
kernel configuration option.
.TP
.IR cpu " (since Linux 2.6.24; " \fBCONFIG_CGROUP_SCHED\fP )
Cgroups can be guaranteed a minimum number of "CPU shares"
when a system is busy.
This does not limit a cgroup's CPU usage if the CPUs are not busy.
For further information, see
.IR Documentation/scheduler/sched-design-CFS.txt .

In Linux 3.2,
this controller was extended to provide CPU "bandwidth" control.
If the kernel is configured with
.BR CONFIG_CFS_BANDWIDTH ,
then within each scheduling period
(defined via a file in the cgroup directory), it is possible to define
an upper limit on the CPU time allocated to the processes in a cgroup.
This upper limit applies even if there is no other competition for the CPU.
Further information can be found in the kernel source file
.IR Documentation/scheduler/sched\-bwc.txt .
.TP
.IR cpuacct " (since Linux 2.6.24; " \fBCONFIG_CGROUP_CPUACCT\fP )
This provides accounting for CPU usage by groups of processes.

Further information can be found in the kernel source file
.IR Documentation/cgroup\-v1/cpuacct.txt .
.TP
.IR cpuset " (since Linux 2.6.24; " \fBCONFIG_CPUSETS\fP )
This cgroup can be used to bind the processes in a cgroup to
a specified set of CPUs and NUMA nodes.

Further information can be found in the kernel source file
.IR Documentation/cgroup\-v1/cpusets.txt .
.TP
.IR memory " (since Linux 2.6.25; " \fBCONFIG_MEMCG\fP )
The memory controller supports reporting and limiting of process memory, kernel
memory, and swap used by cgroups.

Further information can be found in the kernel source file
.IR Documentation/cgroup\-v1/memory.txt .
.TP
.IR devices " (since Linux 2.6.26; " \fBCONFIG_CGROUP_DEVICE\fP )
This supports controlling which processes may create (mknod) devices as
well as open them for reading or writing.
The policies may be specified as whitelists and blacklists.
Hierarchy is enforced, so new rules must not
violate existing rules for the target or ancestor cgroups.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/devices.txt .
.TP
.IR freezer " (since Linux 2.6.28; " \fBCONFIG_CGROUP_FREEZER\fP )
The
.IR freezer
cgroup can suspend and restore (resume) all processes in a cgroup.
Freezing a cgroup
.I /A
also causes its children, for example, processes in
.IR /A/B ,
to be frozen.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/freezer-subsystem.txt .
.TP
.IR net_cls " (since Linux 2.6.29; " \fBCONFIG_CGROUP_NET_CLASSID\fP )
This places a classid, specified for the cgroup, on network packets
created by a cgroup.
These classids can then be used in firewall rules,
as well as used to shape traffic using
.BR tc (8).
This applies only to packets
leaving the cgroup, not to traffic arriving at the cgroup.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/net_cls.txt .
.TP
.IR blkio " (since Linux 2.6.33; " \fBCONFIG_BLK_CGROUP\fP )
The
.I blkio
cgroup controls and limits access to specified block devices by
applying IO control in the form of throttling and upper limits against leaf
nodes and intermediate nodes in the storage hierarchy.

Two policies are available.
The first is a proportional-weight time-based division
of disk implemented with CFQ.
This is in effect for leaf nodes using CFQ.
The second is a throttling policy which specifies
upper I/O rate limits on a device.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/blkio-controller.txt .
.TP
.IR perf_event " (since Linux 2.6.39; " \fBCONFIG_CGROUP_PERF\fP )
This controller allows
.I perf
monitoring of the set of processes grouped in a cgroup.

Further information can be found in the kernel source file
.IR tools/perf/Documentation/perf-record.txt .
.TP
.IR net_prio " (since Linux 3.3; " \fBCONFIG_CGROUP_NET_PRIO\fP )
This allows priorities to be specified, per network interface, for cgroups.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/net_prio.txt .
.TP
.IR hugetlb " (since Linux 3.5; " \fBCONFIG_CGROUP_HUGETLB\fP )
This supports limiting the use of huge pages by cgroups.

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/hugetlb.txt .
.TP
.IR pids " (since Linux 4.3; " \fBCONFIG_CGROUP_PIDS\fP )
This controller permits limiting the number of process that may be created
in a cgroup (and its descendants).

Further information can be found in the kernel source file
.IR Documentation/cgroup-v1/pids.txt .
.\"
.SS Creating cgroups and moving processes
A cgroup filesystem initially contains a single root cgroup, '/',
which all processes belong to.
A new cgroup is created by creating a directory in the cgroup filesystem:

    mkdir /sys/fs/cgroup/cpu/cg1

This creates a new empty cgroup.

A process may be moved to this cgroup by writing its PID into the cgroup's
.I cgroup.procs
file:

    echo $$ > /sys/fs/cgroup/cpu/cg1/cgroup.procs

Only one PID at a time should be written to this file.

Writing the value 0 to a
.IR cgroup.procs
file causes the writing process to be moved to the corresponding cgroup.

When writing a PID into the
.IR cgroup.procs ,
all threads in the process are moved into the new cgroup at once.

Within a hierarchy, a process can be a member of exactly one cgroup.
Writing a process's PID to a
.IR cgroup.procs
file automatically removes it from the cgroup of
which it was previously a member.

The
.I cgroup.procs
file can be read to obtain a list of the processes that are
members of a cgroup.
The returned list of PIDs is not guaranteed to be in order.
Nor is it guaranteed to be free of duplicates.
(For example, a PID may be recycled while reading from the list.)

In cgroups v1 (but not cgroups v2), an individual thread can be moved to
another cgroup by writing its thread ID
(i.e., the kernel thread ID returned by
.BR clone (2)
and
.BR gettid (2))
to the
.IR tasks
file in a cgroup directory.
This file can be read to discover the set of threads
that are members of the cgroup.
This file is not present in cgroup v2 directories.
.\"
.SS Removing cgroups
To remove a cgroup,
it must first have no child cgroups and contain no (nonzombie) processes.
So long as that is the case, one can simply
remove the corresponding directory pathname.
Note that files in a cgroup directory cannot and need not be
removed.
.\"
.SS Cgroups v1 release notification
Two files can be used to determine whether the kernel provides
notifications when a cgroup becomes empty.
A cgroup is considered to be empty when it contains no child
cgroups and no member processes.

A special file in the root directory of each cgroup hierarchy,
.IR release_agent ,
can be used to register the pathname of a program that may be invoked when
a cgroup in the hierarchy becomes empty.
The pathname of the newly empty cgroup (relative to the cgroup mount point)
is provided as the sole command-line argument when the
.IR release_agent
program is invoked.
The
.IR release_agent
program might remove the cgroup directory,
or perhaps repopulate with a process.

The default value of the
.IR release_agent
file is empty, meaning that no release agent is invoked.

Whether or not the
.IR release_agent
program is invoked when a particular cgroup becomes empty is determined
by the value in the
.IR notify_on_release
file in the corresponding cgroup directory.
If this file contains the value 0, then the
.IR release_agent
program is not invoked.
If it contains the value 1, the
.IR release_agent
program is invoked.
The default value for this file in the root cgroup is 0.
At the time when a new cgroup is created,
the value in this file is inherited from the corresponding file
in the parent cgroup.
.\"
.SS Cgroups version 2
In cgroups v2,
all mounted controllers reside in a single unified hierarchy.
While (different) controllers may be simultaneously
mounted under the v1 and v2 hierarchies,
it is not possible to mount the same controller simultaneously
under both the v1 and the v2 hierarchies.

The new behaviors in cgroups v2 are summarized here,
and in some cases elaborated in the following subsections.
.IP 1. 3
Cgroups v2 provides a unified hierarchy against
which all controllers are mounted.
.IP 2.
"Internal" processes are not permitted.
With the exception of the root cgroup, processes may reside
only in leaf nodes (cgroups that do not themselves contain child cgroups).
.IP 3.
Active cgroups must be specified via the files
.IR cgroup.controllers
and
.IR cgroup.subtree_control .
.IP 4.
The
.I tasks
file has been removed.
In addition, the
.I cgroup.clone_children
file that is employed by the
.I cpuset
controller has been removed.
.IP 5.
An improved mechanism for notification of empty cgroups is provided by the
.IR cgroup.events
file.
.PP
For more changes, see the
.I Documentation/cgroup-v2.txt
file in the kernel source.
.\"
.SS Cgroups v2 unified hierarchy
In cgroups v1, the ability to mount different controllers
against different hierarchies was intended to allow great flexibility
for application design.
In practice, though, the flexibility turned out to less useful than expected,
and in many cases added complexity.
Therefore, in cgroups v2,
all available controllers are mounted against a single hierarchy.
The available controllers are automatically mounted,
meaning that it is not necessary (or possible) to specify the controllers
when mounting the cgroup v2 filesystem using a command such as the following:

    mount -t cgroup2 none /mnt/cgroup2

A cgroup v2 controller is available only if it is not currently in use
via a mount against a cgroup v1 hierarchy.
Or, to put things another way, it is not possible to employ
the same controller against both a v1 hierarchy and the unified v2 hierarchy.
.\"
.SS Cgroups v2 """no internal processes""" rule
With the exception of the root cgroup, processes may reside
only in leaf nodes (cgroups that do not themselves contain child cgroups).
This avoids the need to decide how to partition resources between
processes which are members of cgroup A and processes in child cgroups of A.

For instance, if cgroup
.I /cg1/cg2
exists, then a process may reside in
.IR /cg1/cg2 ,
but not in
.IR /cg1 .
This is to avoid an ambiguity in cgroups v1
with respect to the delegation of resources between processes in
.I /cg1
and its child cgroups.
The recommended approach in cgroups v2 is to create a subdirectory called
.I leaf
for any nonleaf cgroup which should contain processes, but no child cgroups.
Thus, processes which previously would have gone into
.I /cg1
would now go into
.IR /cg1/leaf .
This has the advantage of making explicit
the relationship between processes in
.I /cg1/leaf
and
.IR /cg1 's
other children.
.\"
.SS Cgroups v2 subtree control
When a cgroup
.I A/b
is created, its
.IR cgroup.controllers
file contains the list of controllers which were active in its parent, A.
This is the list of controllers which are available to this cgroup.
No controllers are active until they are enabled through the
.IR cgroup.subtree_control
file, by writing the list of space-delimited names of the controllers,
each preceded by '+' (to enable) or '\-' (to disable).
If the
.I freezer
controller is not enabled in
.IR /A/B ,
then it cannot be enabled in
.IR /A/B/C .
.\"
.SS Cgroups v2 cgroup.events file
With cgroups v2, a new mechanism is provided to obtain notification
about when a cgroup becomes empty.
The cgroups v1
.IR release_agent
and
.IR notify_on_release
files are removed, and replaced by a new, more general-purpose file,
.IR cgroup.events .
This file contains key-value pairs
(delimited by newline characters, with the key and value separated by spaces)
that identify events or state for a cgroup.
Currently, only one key appears in this file,
.IR populated ,
which has either the value 0,
meaning that the cgroup (and its descendants)
contain no (nonzombie) processes,
or 1, meaning that the cgroup contains member processes.

The
.IR cgroup.events
file can be monitored, in order to receive notification when a cgroup
transitions between the populated and unpopulated states (or vice versa).
When monitoring this file using
.BR inotify (7),
transitions generate
.BR IN_MODIFY
events, and when monitoring the file using
.BR poll (2),
transitions generate
.B POLLPRI
events.

The cgroups v2
.IR notify_on_release
mechanism offers at least two advantages over the cgroups v1
.IR release_agent
mechanism.
First, it allows for cheaper notification,
since a single process can monitor multiple
.IR cgroup.events
files.
By contrast, the cgroups v1 mechanism requires the creation
of a process for each notification.
Second, notification can be delegated to a process that lives inside
a container associated with the newly empty cgroup.
.\"
.SS /proc files
.TP
.IR /proc/cgroups " (since Linux 2.6.24)"
This file contains information about the controllers
that are compiled into the kernel.
An example of the contents of this file (reformatted for readability)
is the following:

.nf
.in +4n
#subsys_name    hierarchy      num_cgroups    enabled
cpuset          4              1              1
cpu             8              1              1
cpuacct         8              1              1
blkio           6              1              1
memory          3              1              1
devices         10             84             1
freezer         7              1              1
net_cls         9              1              1
perf_event      5              1              1
net_prio        9              1              1
hugetlb         0              1              0
pids            2              1              1
.in
.fi

The fields in this file are, from left to right:
.RS
.IP 1. 3
The name of the controller.
.IP 2.
The unique ID of the cgroup hierarchy on which this controller is mounted.
If multiple cgroups v1 controllers are bound to the same hierarchy,
then each will show the same hierarchy ID in this field.
The value in this field will be 0 if:
.RS 5
.IP a) 3
the controller is not mounted on a cgroups v1 hierarchy;
.IP b)
the controller is bound to the cgroups v2 single unified hierarchy; or
.IP c)
the controller is disabled (see below).
.RE
.IP 3.
The number of control groups in this hierarchy using this controller.
.IP 4.
This field contains the value 1 if this controller is enabled,
or 0 if it has been disabled (via the
.IR cgroup_disable
kernel command-line boot parameter).
.RE
.TP
.IR /proc/[pid]/cgroup " (since Linux 2.6.24)"
This file describes control groups to which the process
with the corresponding PID belongs.
The displayed information differs for
cgroups version 1 and version 2 hierarchies.

For each cgroup hierarchy of which the process is a member,
there is one entry containing three
colon-separated fields of the form:

     hierarchy-ID:controller-list:cgroup-path

For example:
.nf
.ft CW

    5:cpuacct,cpu,cpuset:/daemons
.ft
.fi
.IP
The colon-separated fields are, from left to right:
.RS
.IP 1. 3
For cgroups version 1 hierarchies,
this field contains a unique hierarchy ID number
that can be matched to a hierarchy ID in
.IR /proc/cgroups .
For the cgroups version 2 hierarchy, this field contains the value 0.
.IP 2.
For cgroups version 1 hierarchies,
this field contains a comma-separated list of the controllers
bound to the hierarchy.
For the cgroups version 2 hierarchy, this field is empty.
.IP 3.
This field contains the pathname of the control group in the hierarchy
to which the process belongs.
This pathname is relative to the mount point of the hierarchy.
.RE
.SH ERRORS
The following errors can occur for
.BR mount (2):
.TP
.B EBUSY
An attempt to mount a cgroup version 1 filesystem specified neither the
.I name=
option (to mount a named hierarchy) nor a controller name (or
.IR all ).
.SH NOTES
A child process created via
.BR fork (2)
inherits its parent's cgroup memberships.
A process's cgroup memberships are preserved across
.BR execve (2).
.SH SEE ALSO
.BR prlimit (1),
.BR systemd (1),
.BR clone (2),
.BR ioprio_set (2),
.BR perf_event_open (2),
.BR setrlimit (2),
.BR cgroup_namespaces (7),
.BR cpuset (7),
.BR namespaces (7),
.BR sched (7),
.BR user_namespaces (7)