Control Group v2

Control Group v2

October, 2015 Tejun Heo <tj@kernel.org>

This is the authoritative documentation on the design, interface and

conventions of cgroup v2.  It describes all userland-visible aspects

of cgroup including core and specific controller behaviors.  All

future changes must be reflected in this document.  Documentation for

v1 is available under Documentation/cgroup-v1/.

CONTENTS

1. Introduction

  1-1. Terminology

  1-2. What is cgroup?

2. Basic Operations

  2-1. Mounting

  2-2. Organizing Processes

  2-3. [Un]populated Notification

  2-4. Controlling Controllers

    2-4-1. Enabling and Disabling

    2-4-2. Top-down Constraint

    2-4-3. No Internal Process Constraint

  2-5. Delegation

    2-5-1. Model of Delegation

    2-5-2. Delegation Containment

  2-6. Guidelines

    2-6-1. Organize Once and Control

    2-6-2. Avoid Name Collisions

3. Resource Distribution Models

  3-1. Weights

  3-2. Limits

  3-3. Protections

  3-4. Allocations

4. Interface Files

  4-1. Format

  4-2. Conventions

  4-3. Core Interface Files

5. Controllers

  5-1. CPU

    5-1-1. CPU Interface Files

  5-2. Memory

    5-2-1. Memory Interface Files

    5-2-2. Usage Guidelines

    5-2-3. Memory Ownership

  5-3. IO

    5-3-1. IO Interface Files

    5-3-2. Writeback

6. Namespace

  6-1. Basics

  6-2. The Root and Views

  6-3. Migration and setns(2)

  6-4. Interaction with Other Namespaces

P. Information on Kernel Programming

  P-1. Filesystem Support for Writeback

D. Deprecated v1 Core Features

R. Issues with v1 and Rationales for v2

  R-1. Multiple Hierarchies

  R-2. Thread Granularity

  R-3. Competition Between Inner Nodes and Threads

  R-4. Other Interface Issues

  R-5. Controller Issues and Remedies

    R-5-1. Memory

1. Introduction

1-1. Terminology

"cgroup" stands for "control group" and is never capitalized.  The

singular form is used to designate the whole feature and also as a

qualifier as in "cgroup controllers".  When explicitly referring to

multiple individual control groups, the plural form "cgroups" is used.

1-2. What is cgroup?

cgroup is a mechanism to organize processes hierarchically and

distribute system resources along the hierarchy in a controlled and

configurable manner.

cgroup is largely composed of two parts - the core and controllers.

cgroup core is primarily responsible for hierarchically organizing

processes.  A cgroup controller is usually responsible for

distributing a specific type of system resource along the hierarchy

although there are utility controllers which serve purposes other than

resource distribution.

cgroups form a tree structure and every process in the system belongs

to one and only one cgroup.  All threads of a process belong to the

same cgroup.  On creation, all processes are put in the cgroup that

the parent process belongs to at the time.  A process can be migrated

to another cgroup.  Migration of a process doesn't affect already

existing descendant processes.

Following certain structural constraints, controllers may be enabled or

disabled selectively on a cgroup.  All controller behaviors are

hierarchical - if a controller is enabled on a cgroup, it affects all

processes which belong to the cgroups consisting the inclusive

sub-hierarchy of the cgroup.  When a controller is enabled on a nested

cgroup, it always restricts the resource distribution further.  The

restrictions set closer to the root in the hierarchy can not be

overridden from further away.

2. Basic Operations

2-1. Mounting

Unlike v1, cgroup v2 has only single hierarchy.  The cgroup v2

hierarchy can be mounted with the following mount command.

  # mount -t cgroup2 none $MOUNT_POINT

cgroup2 filesystem has the magic number 0x63677270 ("cgrp").  All

controllers which support v2 and are not bound to a v1 hierarchy are

automatically bound to the v2 hierarchy and show up at the root.

Controllers which are not in active use in the v2 hierarchy can be

bound to other hierarchies.  This allows mixing v2 hierarchy with the

legacy v1 multiple hierarchies in a fully backward compatible way.

A controller can be moved across hierarchies only after the controller

is no longer referenced in its current hierarchy.  Because per-cgroup

controller states are destroyed asynchronously and controllers may

have lingering references, a controller may not show up immediately on

the v2 hierarchy after the final umount of the previous hierarchy.

Similarly, a controller should be fully disabled to be moved out of

the unified hierarchy and it may take some time for the disabled

controller to become available for other hierarchies; furthermore, due

to inter-controller dependencies, other controllers may need to be

disabled too.

While useful for development and manual configurations, moving

controllers dynamically between the v2 and other hierarchies is

strongly discouraged for production use.  It is recommended to decide

the hierarchies and controller associations before starting using the

controllers after system boot.

During transition to v2, system management software might still

automount the v1 cgroup filesystem and so hijack all controllers

during boot, before manual intervention is possible. To make testing

and experimenting easier, the kernel parameter cgroup_no_v1= allows

disabling controllers in v1 and make them always available in v2.

2-2. Organizing Processes

Initially, only the root cgroup exists to which all processes belong.

A child cgroup can be created by creating a sub-directory.

  # mkdir $CGROUP_NAME

A given cgroup may have multiple child cgroups forming a tree

structure.  Each cgroup has a read-writable interface file

"cgroup.procs".  When read, it lists the PIDs of all processes which

belong to the cgroup one-per-line.  The PIDs are not ordered and the

same PID may show up more than once if the process got moved to

another cgroup and then back or the PID got recycled while reading.

A process can be migrated into a cgroup by writing its PID to the

target cgroup's "cgroup.procs" file.  Only one process can be migrated

on a single write(2) call.  If a process is composed of multiple

threads, writing the PID of any thread migrates all threads of the

process.

When a process forks a child process, the new process is born into the

cgroup that the forking process belongs to at the time of the

operation.  After exit, a process stays associated with the cgroup

that it belonged to at the time of exit until it's reaped; however, a

zombie process does not appear in "cgroup.procs" and thus can't be

moved to another cgroup.

A cgroup which doesn't have any children or live processes can be

destroyed by removing the directory.  Note that a cgroup which doesn't

have any children and is associated only with zombie processes is

considered empty and can be removed.

  # rmdir $CGROUP_NAME

"/proc/$PID/cgroup" lists a process's cgroup membership.  If legacy

cgroup is in use in the system, this file may contain multiple lines,

one for each hierarchy.  The entry for cgroup v2 is always in the

format "0::$PATH".

  # cat /proc/842/cgroup

  ...

  0::/test-cgroup/test-cgroup-nested

If the process becomes a zombie and the cgroup it was associated with

is removed subsequently, " (deleted)" is appended to the path.

  # cat /proc/842/cgroup

  ...

  0::/test-cgroup/test-cgroup-nested (deleted)

2-3. [Un]populated Notification

Each non-root cgroup has a "cgroup.events" file which contains

"populated" field indicating whether the cgroup's sub-hierarchy has

live processes in it.  Its value is 0 if there is no live process in

the cgroup and its descendants; otherwise, 1.  poll and [id]notify

events are triggered when the value changes.  This can be used, for

example, to start a clean-up operation after all processes of a given

sub-hierarchy have exited.  The populated state updates and

notifications are recursive.  Consider the following sub-hierarchy

where the numbers in the parentheses represent the numbers of processes

in each cgroup.

  A(4) - B(0) - C(1)

              \ D(0)

A, B and C's "populated" fields would be 1 while D's 0.  After the one

process in C exits, B and C's "populated" fields would flip to "0" and

file modified events will be generated on the "cgroup.events" files of

both cgroups.

2-4. Controlling Controllers

2-4-1. Enabling and Disabling

1a9a5

Each cgroup has a "cgroup.controllers" file which lists all

controllers available for the cgroup to enable.

  # cat cgroup.controllers

  cpu io memory

No controller is enabled by default.  Controllers can be enabled and

disabled by writing to the "cgroup.subtree_control" file.

  # echo "+cpu +memory -io" > cgroup.subtree_control

Only controllers which are listed in "cgroup.controllers" can be

enabled.  When multiple operations are specified as above, either they

all succeed or fail.  If multiple operations on the same controller

are specified, the last one is effective.

Enabling a controller in a cgroup indicates that the distribution of

the target resource across its immediate children will be controlled.

Consider the following sub-hierarchy.  The enabled controllers are

listed in parentheses.

  A(cpu,memory) - B(memory) - C()

                            \ D()

As A has "cpu" and "memory" enabled, A will control the distribution

of CPU cycles and memory to its children, in this case, B.  As B has

"memory" enabled but not "CPU", C and D will compete freely on CPU

cycles but their division of memory available to B will be controlled.

As a controller regulates the distribution of the target resource to

the cgroup's children, enabling it creates the controller's interface

files in the child cgroups.  In the above example, enabling "cpu" on B

would create the "cpu." prefixed controller interface files in C and

D.  Likewise, disabling "memory" from B would remove the "memory."

prefixed controller interface files from C and D.  This means that the

controller interface files - anything which doesn't start with

"cgroup." are owned by the parent rather than the cgroup itself.

2-4-2. Top-down Constraint

Resources are distributed top-down and a cgroup can further distribute

a resource only if the resource has been distributed to it from the

parent.  This means that all non-root "cgroup.subtree_control" files

can only contain controllers which are enabled in the parent's

"cgroup.subtree_control" file.  A controller can be enabled only if

the parent has the controller enabled and a controller can't be

disabled if one or more children have it enabled.

2-4-3. No Internal Process Constraint

Non-root cgroups can only distribute resources to their children when

they don't have any processes of their own.  In other words, only

cgroups which don't contain any processes can have controllers enabled

in their "cgroup.subtree_control" files.

This guarantees that, when a controller is looking at the part of the

hierarchy which has it enabled, processes are always only on the

leaves.  This rules out situations where child cgroups compete against

internal processes of the parent.

The root cgroup is exempt from this restriction.  Root contains

processes and anonymous resource consumption which can't be associated

with any other cgroups and requires special treatment from most

controllers.  How resource consumption in the root cgroup is governed

is up to each controller.

Note that the restriction doesn't get in the way if there is no

enabled controller in the cgroup's "cgroup.subtree_control".  This is

important as otherwise it wouldn't be possible to create children of a

populated cgroup.  To control resource distribution of a cgroup, the

cgroup must create children and transfer all its processes to the

children before enabling controllers in its "cgroup.subtree_control"

file.

2-5. Delegation

2-5-1. Model of Delegation

A cgroup can be delegated to a less privileged user by granting write

access of the directory and its "cgroup.procs" file to the user.  Note

that resource control interface files in a given directory control the

distribution of the parent's resources and thus must not be delegated

along with the directory.

Once delegated, the user can build sub-hierarchy under the directory,

organize processes as it sees fit and further distribute the resources

it received from the parent.  The limits and other settings of all

resource controllers are hierarchical and regardless of what happens

in the delegated sub-hierarchy, nothing can escape the resource

restrictions imposed by the parent.

Currently, cgroup doesn't impose any restrictions on the number of

cgroups in or nesting depth of a delegated sub-hierarchy; however,

this may be limited explicitly in the future.

2-5-2. Delegation Containment

A delegated sub-hierarchy is contained in the sense that processes

can't be moved into or out of the sub-hierarchy by the delegatee.  For

a process with a non-root euid to migrate a target process into a

cgroup by writing its PID to the "cgroup.procs" file, the following

conditions must be met.

- The writer's euid must match either uid or suid of the target process.

- The writer must have write access to the "cgroup.procs" file.

- The writer must have write access to the "cgroup.procs" file of the

  common ancestor of the source and destination cgroups.

The above three constraints ensure that while a delegatee may migrate

processes around freely in the delegated sub-hierarchy it can't pull

in from or push out to outside the sub-hierarchy.

For an example, let's assume cgroups C0 and C1 have been delegated to

user U0 who created C00, C01 under C0 and C10 under C1 as follows and

all processes under C0 and C1 belong to U0.

  ~~~~~~~~~~~~~ - C0 - C00

  ~ cgroup    ~      \ C01

  ~ hierarchy ~

  ~~~~~~~~~~~~~ - C1 - C10

Let's also say U0 wants to write the PID of a process which is

currently in C10 into "C00/cgroup.procs".  U0 has write access to the

file and uid match on the process; however, the common ancestor of the

source cgroup C10 and the destination cgroup C00 is above the points

of delegation and U0 would not have write access to its "cgroup.procs"

files and thus the write will be denied with -EACCES.

2-6. Guidelines

2-6-1. Organize Once and Control

Migrating a process across cgroups is a relatively expensive operation

and stateful resources such as memory are not moved together with the

process.  This is an explicit design decision as there often exist

inherent trade-offs between migration and various hot paths in terms

of synchronization cost.

As such, migrating processes across cgroups frequently as a means to

apply different resource restrictions is discouraged.  A workload

should be assigned to a cgroup according to the system's logical and

resource structure once on start-up.  Dynamic adjustments to resource

distribution can be made by changing controller configuration through

the interface files.

2-6-2. Avoid Name Collisions

Interface files for a cgroup and its children cgroups occupy the same

directory and it is possible to create children cgroups which collide

with interface files.

All cgroup core interface files are prefixed with "cgroup." and each

controller's interface files are prefixed with the controller name and

a dot.  A controller's name is composed of lower case alphabets and

'_'s but never begins with an '_' so it can be used as the prefix

character for collision avoidance.  Also, interface file names won't

start or end with terms which are often used in categorizing workloads

such as job, service, slice, unit or workload.

cgroup doesn't do anything to prevent name collisions and it's the

user's responsibility to avoid them.

3. Resource Distribution Models

cgroup controllers implement several resource distribution schemes

depending on the resource type and expected use cases.  This section

describes major schemes in use along with their expected behaviors.

3-1. Weights

A parent's resource is distributed by adding up the weights of all

active children and giving each the fraction matching the ratio of its

weight against the sum.  As only children which can make use of the

resource at the moment participate in the distribution, this is

work-conserving.  Due to the dynamic nature, this model is usually

used for stateless resources.

All weights are in the range [1, 10000] with the default at 100.  This

allows symmetric multiplicative biases in both directions at fine

enough granularity while staying in the intuitive range.

As long as the weight is in range, all configuration combinations are

valid and there is no reason to reject configuration changes or

process migrations.

"cpu.weight" proportionally distributes CPU cycles to active children

and is an example of this type.

3-2. Limits

A child can only consume upto the configured amount of the resource.

Limits can be over-committed - the sum of the limits of children can

exceed the amount of resource available to the parent.

Limits are in the range [0, max] and defaults to "max", which is noop.

As limits can be over-committed, all configuration combinations are

valid and there is no reason to reject configuration changes or

process migrations.

"io.max" limits the maximum BPS and/or IOPS that a cgroup can consume

on an IO device and is an example of this type.

3-3. Protections

A cgroup is protected to be allocated upto the configured amount of

the resource if the usages of all its ancestors are under their

protected levels.  Protections can be hard guarantees or best effort

soft boundaries.  Protections can also be over-committed in which case

only upto the amount available to the parent is protected among

children.

Protections are in the range [0, max] and defaults to 0, which is

noop.

As protections can be over-committed, all configuration combinations

are valid and there is no reason to reject configuration changes or

process migrations.

"memory.low" implements best-effort memory protection and is an

example of this type.

3-4. Allocations

A cgroup is exclusively allocated a certain amount of a finite

resource.  Allocations can't be over-committed - the sum of the

allocations of children can not exceed the amount of resource

available to the parent.

Allocations are in the range [0, max] and defaults to 0, which is no

resource.

As allocations can't be over-committed, some configuration

combinations are invalid and should be rejected.  Also, if the

resource is mandatory for execution of processes, process migrations

may be rejected.

"cpu.rt.max" hard-allocates realtime slices and is an example of this

type.

4. Interface Files

4-1. Format

All interface files should be in one of the following formats whenever

possible.

  New-line separated values

  (when only one value can be written at once)

VAL0\n
VAL1\n
...

  Space separated values

  (when read-only or multiple values can be written at once)

VAL0 VAL1 ...\n

  Flat keyed

KEY0 VAL0\n
KEY1 VAL1\n
...

  Nested keyed

KEY0 SUB_KEY0=VAL00 SUB_KEY1=VAL01...
KEY1 SUB_KEY0=VAL10 SUB_KEY1=VAL11...
...

For a writable file, the format for writing should generally match

reading; however, controllers may allow omitting later fields or

implement restricted shortcuts for most common use cases.

For both flat and nested keyed files, only the values for a single key

can be written at a time.  For nested keyed files, the sub key pairs

may be specified in any order and not all pairs have to be specified.

4-2. Conventions

- Settings for a single feature should be contained in a single file.

- The root cgroup should be exempt from resource control and thus

  shouldn't have resource control interface files.  Also,

  informational files on the root cgroup which end up showing global

  information available elsewhere shouldn't exist.

- If a controller implements weight based resource distribution, its

  interface file should be named "weight" and have the range [1,

  10000] with 100 as the default.  The values are chosen to allow

  enough and symmetric bias in both directions while keeping it

  intuitive (the default is 100%).

- If a controller implements an absolute resource guarantee and/or

  limit, the interface files should be named "min" and "max"

  respectively.  If a controller implements best effort resource

  guarantee and/or limit, the interface files should be named "low"

  and "high" respectively.

  In the above four control files, the special token "max" should be

  used to represent upward infinity for both reading and writing.

- If a setting has a configurable default value and keyed specific

  overrides, the default entry should be keyed with "default" and

  appear as the first entry in the file.

  The default value can be updated by writing either "default $VAL" or

  "$VAL".

  When writing to update a specific override, "default" can be used as

  the value to indicate removal of the override.  Override entries

  with "default" as the value must not appear when read.

  For example, a setting which is keyed by major:minor device numbers

  with integer values may look like the following.

    # cat cgroup-example-interface-file

    default 150

    8:0 300

  The default value can be updated by

    # echo 125 > cgroup-example-interface-file

  or

    # echo "default 125" > cgroup-example-interface-file

  An override can be set by

    # echo "8:16 170" > cgroup-example-interface-file

  and cleared by

    # echo "8:0 default" > cgroup-example-interface-file

    # cat cgroup-example-interface-file

    default 125

    8:16 170

- For events which are not very high frequency, an interface file

  "events" should be created which lists event key value pairs.

  Whenever a notifiable event happens, file modified event should be

  generated on the file.

4-3. Core Interface Files

All cgroup core files are prefixed with "cgroup."

  cgroup.procs

A read-write new-line separated values file which exists on
all cgroups.

When read, it lists the PIDs of all processes which belong to
the cgroup one-per-line.  The PIDs are not ordered and the
same PID may show up more than once if the process got moved
to another cgroup and then back or the PID got recycled while
reading.

A PID can be written to migrate the process associated with
the PID to the cgroup.  The writer should match all of the
following conditions.

- Its euid is either root or must match either uid or suid of

          the target process.

- It must have write access to the "cgroup.procs" file.

- It must have write access to the "cgroup.procs" file of the
 common ancestor of the source and destination cgroups.

When delegating a sub-hierarchy, write access to this file
should be granted along with the containing directory.

  cgroup.controllers

A read-only space separated values file which exists on all
cgroups.

It shows space separated list of all controllers available to
the cgroup.  The controllers are not ordered.

  cgroup.subtree_control

A read-write space separated values file which exists on all
cgroups.  Starts out empty.

When read, it shows space separated list of the controllers
which are enabled to control resource distribution from the
cgroup to its children.

Space separated list of controllers prefixed with '+' or '-'
can be written to enable or disable controllers.  A controller
name prefixed with '+' enables the controller and '-'
disables.  If a controller appears more than once on the list,
the last one is effective.  When multiple enable and disable
operations are specified, either all succeed or all fail.

  cgroup.events

A read-only flat-keyed file which exists on non-root cgroups.
The following entries are defined.  Unless specified
otherwise, a value change in this file generates a file
modified event.

 populated

1 if the cgroup or its descendants contains any live
processes; otherwise, 0.

5. Controllers

5-1. CPU

[NOTE: The interface for the cpu controller hasn't been merged yet]

The "cpu" controllers regulates distribution of CPU cycles.  This

controller implements weight and absolute bandwidth limit models for

normal scheduling policy and absolute bandwidth allocation model for

realtime scheduling policy.

5-1-1. CPU Interface Files

All time durations are in microseconds.

  cpu.stat

A read-only flat-keyed file which exists on non-root cgroups.

It reports the following six stats.

 usage_usec
 user_usec
 system_usec
 nr_periods
 nr_throttled
 throttled_usec

  cpu.weight

A read-write single value file which exists on non-root
cgroups.  The default is "100".

The weight in the range [1, 10000].

  cpu.max

A read-write two value file which exists on non-root cgroups.
The default is "max 100000".

The maximum bandwidth limit.  It's in the following format.

 $MAX $PERIOD

which indicates that the group may consume upto $MAX in each
$PERIOD duration.  "max" for $MAX indicates no limit.  If only
one number is written, $MAX is updated.

  cpu.rt.max

  [NOTE: The semantics of this file is still under discussion and the

   interface hasn't been merged yet]

A read-write two value file which exists on all cgroups.
The default is "0 100000".

The maximum realtime runtime allocation.  Over-committing
configurations are disallowed and process migrations are
rejected if not enough bandwidth is available.  It's in the
following format.

 $MAX $PERIOD

which indicates that the group may consume upto $MAX in each
$PERIOD duration.  If only one number is written, $MAX is
updated.

5-2. Memory

The "memory" controller regulates distribution of memory.  Memory is

stateful and implements both limit and protection models.  Due to the

intertwining between memory usage and reclaim pressure and the

stateful nature of memory, the distribution model is relatively

complex.

While not completely water-tight, all major memory usages by a given

cgroup are tracked so that the total memory consumption can be

accounted and controlled to a reasonable extent.  Currently, the

following types of memory usages are tracked.

- Userland memory - page cache and anonymous memory.

- Kernel data structures such as dentries and inodes.

- TCP socket buffers.

The above list may expand in the future for better coverage.

5-2-1. Memory Interface Files

All memory amounts are in bytes.  If a value which is not aligned to

PAGE_SIZE is written, the value may be rounded up to the closest

PAGE_SIZE multiple when read back.

  memory.current

A read-only single value file which exists on non-root
cgroups.

The total amount of memory currently being used by the cgroup
and its descendants.

  memory.low

A read-write single value file which exists on non-root
cgroups.  The default is "0".

Best-effort memory protection.  If the memory usages of a
cgroup and all its ancestors are below their low boundaries,
the cgroup's memory won't be reclaimed unless memory can be
reclaimed from unprotected cgroups.

Putting more memory than generally available under this
protection is discouraged.

  memory.high

A read-write single value file which exists on non-root
cgroups.  The default is "max".

Memory usage throttle limit.  This is the main mechanism to
control memory usage of a cgroup.  If a cgroup's usage goes
over the high boundary, the processes of the cgroup are
throttled and put under heavy reclaim pressure.

Going over the high limit never invokes the OOM killer and
under extreme conditions the limit may be breached.

  memory.max

A read-write single value file which exists on non-root
cgroups.  The default is "max".

Memory usage hard limit.  This is the final protection
mechanism.  If a cgroup's memory usage reaches this limit and
can't be reduced, the OOM killer is invoked in the cgroup.
Under certain circumstances, the usage may go over the limit
temporarily.

This is the ultimate protection mechanism.  As long as the
high limit is used and monitored properly, this limit's
utility is limited to providing the final safety net.

  memory.events

A read-only flat-keyed file which exists on non-root cgroups.
The following entries are defined.  Unless specified
otherwise, a value change in this file generates a file
modified event.

 low

The number of times the cgroup is reclaimed due to
high memory pressure even though its usage is under
the low boundary.  This usually indicates that the low
boundary is over-committed.

 high

The number of times processes of the cgroup are
throttled and routed to perform direct memory reclaim
because the high memory boundary was exceeded.  For a
cgroup whose memory usage is capped by the high limit
rather than global memory pressure, this event's
occurrences are expected.

 max

The number of times the cgroup's memory usage was
about to go over the max boundary.  If direct reclaim
fails to bring it down, the OOM killer is invoked.

 oom

The number of times the OOM killer has been invoked in
the cgroup.  This may not exactly match the number of
processes killed but should generally be close.

  memory.stat

A read-only flat-keyed file which exists on non-root cgroups.

This breaks down the cgroup's memory footprint into different
types of memory, type-specific details, and other information
on the state and past events of the memory management system.

All memory amounts are in bytes.

The entries are ordered to be human readable, and new entries
can show up in the middle. Don't rely on items remaining in a
fixed position; use the keys to look up specific values!

 anon

Amount of memory used in anonymous mappings such as
brk(), sbrk(), and mmap(MAP_ANONYMOUS)

 file

Amount of memory used to cache filesystem data,
including tmpfs and shared memory.

 kernel_stack

Amount of memory allocated to kernel stacks.

 slab

Amount of memory used for storing in-kernel data
structures.

 sock

Amount of memory used in network transmission buffers

 file_mapped

Amount of cached filesystem data mapped with mmap()

 file_dirty

Amount of cached filesystem data that was modified but
not yet written back to disk

 file_writeback

Amount of cached filesystem data that was modified and
is currently being written back to disk

 inactive_anon
 active_anon
 inactive_file
 active_file
 unevictable

Amount of memory, swap-backed and filesystem-backed,
on the internal memory management lists used by the
page reclaim algorithm

 slab_reclaimable

Part of "slab" that might be reclaimed, such as
dentries and inodes.

 slab_unreclaimable

Part of "slab" that cannot be reclaimed on memory
pressure.

 pgfault

Total number of page faults incurred

 pgmajfault

Number of major page faults incurred

  memory.swap.current

A read-only single value file which exists on non-root
cgroups.

The total amount of swap currently being used by the cgroup
and its descendants.

  memory.swap.max

A read-write single value file which exists on non-root
cgroups.  The default is "max".

Swap usage hard limit.  If a cgroup's swap usage reaches this
limit, anonymous meomry of the cgroup will not be swapped out.

5-2-2. Usage Guidelines

"memory.high" is the main mechanism to control memory usage.

Over-committing on high limit (sum of high limits > available memory)

and letting global memory pressure to distribute memory according to

usage is a viable strategy.

Because breach of the high limit doesn't trigger the OOM killer but

throttles the offending cgroup, a management agent has ample

opportunities to monitor and take appropriate actions such as granting

more memory or terminating the workload.

Determining whether a cgroup has enough memory is not trivial as

memory usage doesn't indicate whether the workload can benefit from

more memory.  For example, a workload which writes data received from

network to a file can use all available memory but can also operate as

performant with a small amount of memory.  A measure of memory

pressure - how much the workload is being impacted due to lack of

memory - is necessary to determine whether a workload needs more

memory; unfortunately, memory pressure monitoring mechanism isn't

implemented yet.

5-2-3. Memory Ownership

A memory area is charged to the cgroup which instantiated it and stays

charged to the cgroup until the area is released.  Migrating a process

to a different cgroup doesn't move the memory usages that it

instantiated while in the previous cgroup to the new cgroup.

A memory area may be used by processes belonging to different cgroups.

To which cgroup the area will be charged is in-deterministic; however,

over time, the memory area is likely to end up in a cgroup which has

enough memory allowance to avoid high reclaim pressure.

If a cgroup sweeps a considerable amount of memory which is expected

to be accessed repeatedly by other cgroups, it may make sense to use

POSIX_FADV_DONTNEED to relinquish the ownership of memory areas

belonging to the affected files to ensure correct memory ownership.

5-3. IO

The "io" controller regulates the distribution of IO resources.  This

controller implements both weight based and absolute bandwidth or IOPS

limit distribution; however, weight based distribution is available

only if cfq-iosched is in use and neither scheme is available for

blk-mq devices.

5-3-1. IO Interface Files

  io.stat

A read-only nested-keyed file which exists on non-root
cgroups.

Lines are keyed by $MAJ:$MIN device numbers and not ordered.
The following nested keys are defined.

 rbytes
Bytes read
 wbytes
Bytes written
 rios Number of read IOs
 wios Number of write IOs

An example read output follows.

 8:16 rbytes=1459200 wbytes=314773504 rios=192 wios=353
 8:0 rbytes=90430464 wbytes=299008000 rios=8950 wios=1252

  io.weight

A read-write flat-keyed file which exists on non-root cgroups.
The default is "default 100".

The first line is the default weight applied to devices
without specific override.  The rest are overrides keyed by
$MAJ:$MIN device numbers and not ordered.  The weights are in
the range [1, 10000] and specifies the relative amount IO time
the cgroup can use in relation to its siblings.

The default weight can be updated by writing either "default
$WEIGHT" or simply "$WEIGHT".  Overrides can be set by writing
"$MAJ:$MIN $WEIGHT" and unset by writing "$MAJ:$MIN default".

An example read output follows.

 default 100
 8:16 200
 8:0 50

  io.max

A read-write nested-keyed file which exists on non-root
cgroups.

BPS and IOPS based IO limit.  Lines are keyed by $MAJ:$MIN
device numbers and not ordered.  The following nested keys are
defined.

 rbps Max read bytes per second
 wbps Max write bytes per second
 riops
Max read IO operations per second
 wiops
Max write IO operations per second

When writing, any number of nested key-value pairs can be
specified in any order.  "max" can be specified as the value
to remove a specific limit.  If the same key is specified
multiple times, the outcome is undefined.

BPS and IOPS are measured in each IO direction and IOs are
delayed if limit is reached.  Temporary bursts are allowed.

Setting read limit at 2M BPS and write at 120 IOPS for 8:16.

 echo "8:16 rbps=2097152 wiops=120" > io.max

Reading returns the following.

 8:16 rbps=2097152 wbps=max riops=max wiops=120

Write IOPS limit can be removed by writing the following.

 echo "8:16 wiops=max" > io.max

Reading now returns the following.

 8:16 rbps=2097152 wbps=max riops=max wiops=max

5-3-2. Writeback

Page cache is dirtied through buffered writes and shared mmaps and

written asynchronously to the backing filesystem by the writeback

mechanism.  Writeback sits between the memory and IO domains and

regulates the proportion of dirty memory by balancing dirtying and

write IOs.

The io controller, in conjunction with the memory controller,

implements control of page cache writeback IOs.  The memory controller

defines the memory domain that dirty memory ratio is calculated and

maintained for and the io controller defines the io domain which

writes out dirty pages for the memory domain.  Both system-wide and

per-cgroup dirty memory states are examined and the more restrictive

of the two is enforced.

cgroup writeback requires explicit support from the underlying

filesystem.  Currently, cgroup writeback is implemented on ext2, ext4

and btrfs.  On other filesystems, all writeback IOs are attributed to

the root cgroup.

There are inherent differences in memory and writeback management

which affects how cgroup ownership is tracked.  Memory is tracked per

page while writeback per inode.  For the purpose of writeback, an

inode is assigned to a cgroup and all IO requests to write dirty pages

from the inode are attributed to that cgroup.

As cgroup ownership for memory is tracked per page, there can be pages

which are associated with different cgroups than the one the inode is

associated with.  These are called foreign pages.  The writeback

constantly keeps track of foreign pages and, if a particular foreign

cgroup becomes the majority over a certain period of time, switches

the ownership of the inode to that cgroup.

While this model is enough for most use cases where a given inode is

mostly dirtied by a single cgroup even when the main writing cgroup

changes over time, use cases where multiple cgroups write to a single

inode simultaneously are not supported well.  In such circumstances, a

significant portion of IOs are likely to be attributed incorrectly.

As memory controller assigns page ownership on the first use and

doesn't update it until the page is released, even if writeback

strictly follows page ownership, multiple cgroups dirtying overlapping

areas wouldn't work as expected.  It's recommended to avoid such usage

patterns.

The sysctl knobs which affect writeback behavior are applied to cgroup

writeback as follows.

  vm.dirty_background_ratio

  vm.dirty_ratio

These ratios apply the same to cgroup writeback with the
amount of available memory capped by limits imposed by the
memory controller and system-wide clean memory.

  vm.dirty_background_bytes

  vm.dirty_bytes

For cgroup writeback, this is calculated into ratio against
total available memory and applied the same way as
vm.dirty[_background]_ratio.

6. Namespace

6-1. Basics

cgroup namespace provides a mechanism to virtualize the view of the

"/proc/$PID/cgroup" file and cgroup mounts.  The CLONE_NEWCGROUP clone

flag can be used with clone(2) and unshare(2) to create a new cgroup

namespace.  The process running inside the cgroup namespace will have

its "/proc/$PID/cgroup" output restricted to cgroupns root.  The

cgroupns root is the cgroup of the process at the time of creation of

the cgroup namespace.

Without cgroup namespace, the "/proc/$PID/cgroup" file shows the

complete path of the cgroup of a process.  In a container setup where

a set of cgroups and namespaces are intended to isolate processes the

"/proc/$PID/cgroup" file may leak potential system level information

to the isolated processes.  For Example:

  # cat /proc/self/cgroup

  0::/batchjobs/container_id1

The path '/batchjobs/container_id1' can be considered as system-data

and undesirable to expose to the isolated processes.  cgroup namespace

can be used to restrict visibility of this path.  For example, before

creating a cgroup namespace, one would see:

  # ls -l /proc/self/ns/cgroup

  lrwxrwxrwx 1 root root 0 2014-07-15 10:37 /proc/self/ns/cgroup -> cgroup:[4026531835]

  # cat /proc/self/cgroup

  0::/batchjobs/container_id1

After unsharing a new namespace, the view changes.

  # ls -l /proc/self/ns/cgroup

  lrwxrwxrwx 1 root root 0 2014-07-15 10:35 /proc/self/ns/cgroup -> cgroup:[4026532183]

  # cat /proc/self/cgroup

  0::/

When some thread from a multi-threaded process unshares its cgroup

namespace, the new cgroupns gets applied to the entire process (all

the threads).  This is natural for the v2 hierarchy; however, for the

legacy hierarchies, this may be unexpected.

A cgroup namespace is alive as long as there are processes inside or

mounts pinning it.  When the last usage goes away, the cgroup

namespace is destroyed.  The cgroupns root and the actual cgroups

remain.

6-2. The Root and Views

The 'cgroupns root' for a cgroup namespace is the cgroup in which the

process calling unshare(2) is running.  For example, if a process in

/batchjobs/container_id1 cgroup calls unshare, cgroup

/batchjobs/container_id1 becomes the cgroupns root.  For the

init_cgroup_ns, this is the real root ('/') cgroup.

The cgroupns root cgroup does not change even if the namespace creator

process later moves to a different cgroup.

  # ~/unshare -c # unshare cgroupns in some cgroup

  # cat /proc/self/cgroup

  0::/

  # mkdir sub_cgrp_1

  # echo 0 > sub_cgrp_1/cgroup.procs

  # cat /proc/self/cgroup

  0::/sub_cgrp_1

Each process gets its namespace-specific view of "/proc/$PID/cgroup"

Processes running inside the cgroup namespace will be able to see

cgroup paths (in /proc/self/cgroup) only inside their root cgroup.

From within an unshared cgroupns:

  # sleep 100000 &

  [1] 7353

  # echo 7353 > sub_cgrp_1/cgroup.procs

  # cat /proc/7353/cgroup

  0::/sub_cgrp_1

From the initial cgroup namespace, the real cgroup path will be

visible:

  $ cat /proc/7353/cgroup

  0::/batchjobs/container_id1/sub_cgrp_1

From a sibling cgroup namespace (that is, a namespace rooted at a

different cgroup), the cgroup path relative to its own cgroup

namespace root will be shown.  For instance, if PID 7353's cgroup

namespace root is at '/batchjobs/container_id2', then it will see

  # cat /proc/7353/cgroup

  0::/../container_id2/sub_cgrp_1

Note that the relative path always starts with '/' to indicate that

its relative to the cgroup namespace root of the caller.

6-3. Migration and setns(2)

Processes inside a cgroup namespace can move into and out of the

namespace root if they have proper access to external cgroups.  For

example, from inside a namespace with cgroupns root at

/batchjobs/container_id1, and assuming that the global hierarchy is

still accessible inside cgroupns:

  # cat /proc/7353/cgroup

  0::/sub_cgrp_1

  # echo 7353 > batchjobs/container_id2/cgroup.procs

  # cat /proc/7353/cgroup

  0::/../container_id2

Note that this kind of setup is not encouraged.  A task inside cgroup

namespace should only be exposed to its own cgroupns hierarchy.

setns(2) to another cgroup namespace is allowed when:

(a) the process has CAP_SYS_ADMIN against its current user namespace

(b) the process has CAP_SYS_ADMIN against the target cgroup

    namespace's userns

No implicit cgroup changes happen with attaching to another cgroup

namespace.  It is expected that the someone moves the attaching

process under the target cgroup namespace root.

6-4. Interaction with Other Namespaces

Namespace specific cgroup hierarchy can be mounted by a process

running inside a non-init cgroup namespace.

  # mount -t cgroup2 none $MOUNT_POINT

This will mount the unified cgroup hierarchy with cgroupns root as the

filesystem root.  The process needs CAP_SYS_ADMIN against its user and

mount namespaces.

The virtualization of /proc/self/cgroup file combined with restricting

the view of cgroup hierarchy by namespace-private cgroupfs mount

provides a properly isolated cgroup view inside the container.

P. Information on Kernel Programming

This section contains kernel programming information in the areas

where interacting with cgroup is necessary.  cgroup core and

controllers are not covered.

P-1. Filesystem Support for Writeback

A filesystem can support cgroup writeback by updating

address_space_operations->writepage[s]() to annotate bio's using the

following two functions.

  wbc_init_bio(@wbc, @bio)

Should be called for each bio carrying writeback data and
associates the bio with the inode's owner cgroup.  Can be
called anytime between bio allocation and submission.

  wbc_account_io(@wbc, @page, @bytes)

Should be called for each data segment being written out.
While this function doesn't care exactly when it's called
during the writeback session, it's the easiest and most
natural to call it as data segments are added to a bio.

With writeback bio's annotated, cgroup support can be enabled per

super_block by setting SB_I_CGROUPWB in ->s_iflags.  This allows for

selective disabling of cgroup writeback support which is helpful when

certain filesystem features, e.g. journaled data mode, are

incompatible.

wbc_init_bio() binds the specified bio to its cgroup.  Depending on

the configuration, the bio may be executed at a lower priority and if

the writeback session is holding shared resources, e.g. a journal

entry, may lead to priority inversion.  There is no one easy solution

for the problem.  Filesystems can try to work around specific problem

cases by skipping wbc_init_bio() or using bio_associate_blkcg()

directly.

D. Deprecated v1 Core Features

- Multiple hierarchies including named ones are not supported.

- All mount options and remounting are not supported.

- The "tasks" file is removed and "cgroup.procs" is not sorted.

- "cgroup.clone_children" is removed.

- /proc/cgroups is meaningless for v2.  Use "cgroup.controllers" file

  at the root instead.

R. Issues with v1 and Rationales for v2

R-1. Multiple Hierarchies

cgroup v1 allowed an arbitrary number of hierarchies and each

hierarchy could host any number of controllers.  While this seemed to

provide a high level of flexibility, it wasn't useful in practice.

For example, as there is only one instance of each controller, utility

type controllers such as freezer which can be useful in all

hierarchies could only be used in one.  The issue is exacerbated by

the fact that controllers couldn't be moved to another hierarchy once

hierarchies were populated.  Another issue was that all controllers

bound to a hierarchy were forced to have exactly the same view of the

hierarchy.  It wasn't possible to vary the granularity depending on

the specific controller.

In practice, these issues heavily limited which controllers could be

put on the same hierarchy and most configurations resorted to putting

each controller on its own hierarchy.  Only closely related ones, such

as the cpu and cpuacct controllers, made sense to be put on the same

hierarchy.  This often meant that userland ended up managing multiple

similar hierarchies repeating the same steps on each hierarchy

whenever a hierarchy management operation was necessary.

Furthermore, support for multiple hierarchies came at a steep cost.

It greatly complicated cgroup core implementation but more importantly

the support for multiple hierarchies restricted how cgroup could be

used in general and what controllers was able to do.

There was no limit on how many hierarchies there might be, which meant

that a thread's cgroup membership couldn't be described in finite

length.  The key might contain any number of entries and was unlimited

in length, which made it highly awkward to manipulate and led to

addition of controllers which existed only to identify membership,

which in turn exacerbated the original problem of proliferating number

of hierarchies.

Also, as a controller couldn't have any expectation regarding the

topologies of hierarchies other controllers might be on, each

controller had to assume that all other controllers were attached to

completely orthogonal hierarchies.  This made it impossible, or at

least very cumbersome, for controllers to cooperate with each other.

In most use cases, putting controllers on hierarchies which are

completely orthogonal to each other isn't necessary.  What usually is

called for is the ability to have differing levels of granularity

depending on the specific controller.  In other words, hierarchy may

be collapsed from leaf towards root when viewed from specific

controllers.  For example, a given configuration might not care about

how memory is distributed beyond a certain level while still wanting

to control how CPU cycles are distributed.

R-2. Thread Granularity

cgroup v1 allowed threads of a process to belong to different cgroups.

This didn't make sense for some controllers and those controllers

ended up implementing different ways to ignore such situations but

much more importantly it blurred the line between API exposed to

individual applications and system management interface.

Generally, in-process knowledge is available only to the process

itself; thus, unlike service-level organization of processes,

categorizing threads of a process requires active participation from

the application which owns the target process.

cgroup v1 had an ambiguously defined delegation model which got abused

in combination with thread granularity.  cgroups were delegated to

individual applications so that they can create and manage their own

sub-hierarchies and control resource distributions along them.  This

effectively raised cgroup to the status of a syscall-like API exposed

to lay programs.

First of all, cgroup has a fundamentally inadequate interface to be

exposed this way.  For a process to access its own knobs, it has to

extract the path on the target hierarchy from /proc/self/cgroup,

construct the path by appending the name of the knob to the path, open

and then read and/or write to it.  This is not only extremely clunky

and unusual but also inherently racy.  There is no conventional way to

define transaction across the required steps and nothing can guarantee

that the process would actually be operating on its own sub-hierarchy.

cgroup controllers implemented a number of knobs which would never be

accepted as public APIs because they were just adding control knobs to

system-management pseudo filesystem.  cgroup ended up with interface

knobs which were not properly abstracted or refined and directly

revealed kernel internal details.  These knobs got exposed to

individual applications through the ill-defined delegation mechanism

effectively abusing cgroup as a shortcut to implementing public APIs

without going through the required scrutiny.

This was painful for both userland and kernel.  Userland ended up with

misbehaving and poorly abstracted interfaces and kernel exposing and

locked into constructs inadvertently.

R-3. Competition Between Inner Nodes and Threads

cgroup v1 allowed threads to be in any cgroups which created an

interesting problem where threads belonging to a parent cgroup and its

children cgroups competed for resources.  This was nasty as two

different types of entities competed and there was no obvious way to

settle it.  Different controllers did different things.

The cpu controller considered threads and cgroups as equivalents and

mapped nice levels to cgroup weights.  This worked for some cases but

fell flat when children wanted to be allocated specific ratios of CPU

cycles and the number of internal threads fluctuated - the ratios

constantly changed as the number of competing entities fluctuated.

There also were other issues.  The mapping from nice level to weight

wasn't obvious or universal, and there were various other knobs which

simply weren't available for threads.

The io controller implicitly created a hidden leaf node for each

cgroup to host the threads.  The hidden leaf had its own copies of all

the knobs with "leaf_" prefixed.  While this allowed equivalent

control over internal threads, it was with serious drawbacks.  It

always added an extra layer of nesting which wouldn't be necessary

otherwise, made the interface messy and significantly complicated the

implementation.

The memory controller didn't have a way to control what happened

between internal tasks and child cgroups and the behavior was not

clearly defined.  There were attempts to add ad-hoc behaviors and

knobs to tailor the behavior to specific workloads which would have

led to problems extremely difficult to resolve in the long term.

Multiple controllers struggled with internal tasks and came up with

different ways to deal with it; unfortunately, all the approaches were

severely flawed and, furthermore, the widely different behaviors

made cgroup as a whole highly inconsistent.

This clearly is a problem which needs to be addressed from cgroup core

in a uniform way.

R-4. Other Interface Issues

cgroup v1 grew without oversight and developed a large number of

idiosyncrasies and inconsistencies.  One issue on the cgroup core side

was how an empty cgroup was notified - a userland helper binary was

forked and executed for each event.  The event delivery wasn't

recursive or delegatable.  The limitations of the mechanism also led

to in-kernel event delivery filtering mechanism further complicating

the interface.

Controller interfaces were problematic too.  An extreme example is

controllers completely ignoring hierarchical organization and treating

all cgroups as if they were all located directly under the root

cgroup.  Some controllers exposed a large amount of inconsistent

implementation details to userland.

There also was no consistency across controllers.  When a new cgroup

was created, some controllers defaulted to not imposing extra

restrictions while others disallowed any resource usage until

explicitly configured.  Configuration knobs for the same type of

control used widely differing naming schemes and formats.  Statistics

and information knobs were named arbitrarily and used different

formats and units even in the same controller.

cgroup v2 establishes common conventions where appropriate and updates

controllers so that they expose minimal and consistent interfaces.

R-5. Controller Issues and Remedies

R-5-1. Memory

The original lower boundary, the soft limit, is defined as a limit

that is per default unset.  As a result, the set of cgroups that

global reclaim prefers is opt-in, rather than opt-out.  The costs for

optimizing these mostly negative lookups are so high that the

implementation, despite its enormous size, does not even provide the

basic desirable behavior.  First off, the soft limit has no

hierarchical meaning.  All configured groups are organized in a global

rbtree and treated like equal peers, regardless where they are located

in the hierarchy.  This makes subtree delegation impossible.  Second,

the soft limit reclaim pass is so aggressive that it not just

introduces high allocation latencies into the system, but also impacts

system performance due to overreclaim, to the point where the feature

becomes self-defeating.

The memory.low boundary on the other hand is a top-down allocated

reserve.  A cgroup enjoys reclaim protection when it and all its

ancestors are below their low boundaries, which makes delegation of

subtrees possible.  Secondly, new cgroups have no reserve per default

and in the common case most cgroups are eligible for the preferred

reclaim pass.  This allows the new low boundary to be efficiently

implemented with just a minor addition to the generic reclaim code,

without the need for out-of-band data structures and reclaim passes.

Because the generic reclaim code considers all cgroups except for the

ones running low in the preferred first reclaim pass, overreclaim of

individual groups is eliminated as well, resulting in much better

overall workload performance.

The original high boundary, the hard limit, is defined as a strict

limit that can not budge, even if the OOM killer has to be called.

But this generally goes against the goal of making the most out of the

available memory.  The memory consumption of workloads varies during

runtime, and that requires users to overcommit.  But doing that with a

strict upper limit requires either a fairly accurate prediction of the

working set size or adding slack to the limit.  Since working set size

estimation is hard and error prone, and getting it wrong results in

OOM kills, most users tend to err on the side of a looser limit and

end up wasting precious resources.

The memory.high boundary on the other hand can be set much more

conservatively.  When hit, it throttles allocations by forcing them

into direct reclaim to work off the excess, but it never invokes the

OOM killer.  As a result, a high boundary that is chosen too

aggressively will not terminate the processes, but instead it will

lead to gradual performance degradation.  The user can monitor this

and make corrections until the minimal memory footprint that still

gives acceptable performance is found.

In extreme cases, with many concurrent allocations and a complete

breakdown of reclaim progress within the group, the high boundary can

be exceeded.  But even then it's mostly better to satisfy the

allocation from the slack available in other groups or the rest of the

system than killing the group.  Otherwise, memory.max is there to

limit this type of spillover and ultimately contain buggy or even

malicious applications.

Setting the original memory.limit_in_bytes below the current usage was

subject to a race condition, where concurrent charges could cause the

limit setting to fail. memory.max on the other hand will first set the

limit to prevent new charges, and then reclaim and OOM kill until the

new limit is met - or the task writing to memory.max is killed.

The combined memory+swap accounting and limiting is replaced by real

control over swap space.

The main argument for a combined memory+swap facility in the original

cgroup design was that global or parental pressure would always be

able to swap all anonymous memory of a child group, regardless of the

child's own (possibly untrusted) configuration.  However, untrusted

groups can sabotage swapping by other means - such as referencing its

anonymous memory in a tight loop - and an admin can not assume full

swappability when overcommitting untrusted jobs.

For trusted jobs, on the other hand, a combined counter is not an

intuitive userspace interface, and it flies in the face of the idea

that cgroup controllers should account and limit specific physical

resources.  Swap space is a resource like all others in the system,

and that's why unified hierarchy allows distributing it separately.