xref: /linux/Documentation/admin-guide/cgroup-v2.rst (revision 2b64b2ed277ff23e785fbdb65098ee7e1252d64f)
1================
2Control Group v2
3================
4
5:Date: October, 2015
6:Author: Tejun Heo <tj@kernel.org>
7
8This is the authoritative documentation on the design, interface and
9conventions of cgroup v2.  It describes all userland-visible aspects
10of cgroup including core and specific controller behaviors.  All
11future changes must be reflected in this document.  Documentation for
12v1 is available under Documentation/cgroup-v1/.
13
14.. CONTENTS
15
16   1. Introduction
17     1-1. Terminology
18     1-2. What is cgroup?
19   2. Basic Operations
20     2-1. Mounting
21     2-2. Organizing Processes and Threads
22       2-2-1. Processes
23       2-2-2. Threads
24     2-3. [Un]populated Notification
25     2-4. Controlling Controllers
26       2-4-1. Enabling and Disabling
27       2-4-2. Top-down Constraint
28       2-4-3. No Internal Process Constraint
29     2-5. Delegation
30       2-5-1. Model of Delegation
31       2-5-2. Delegation Containment
32     2-6. Guidelines
33       2-6-1. Organize Once and Control
34       2-6-2. Avoid Name Collisions
35   3. Resource Distribution Models
36     3-1. Weights
37     3-2. Limits
38     3-3. Protections
39     3-4. Allocations
40   4. Interface Files
41     4-1. Format
42     4-2. Conventions
43     4-3. Core Interface Files
44   5. Controllers
45     5-1. CPU
46       5-1-1. CPU Interface Files
47     5-2. Memory
48       5-2-1. Memory Interface Files
49       5-2-2. Usage Guidelines
50       5-2-3. Memory Ownership
51     5-3. IO
52       5-3-1. IO Interface Files
53       5-3-2. Writeback
54       5-3-3. IO Latency
55         5-3-3-1. How IO Latency Throttling Works
56         5-3-3-2. IO Latency Interface Files
57     5-4. PID
58       5-4-1. PID Interface Files
59     5-5. Cpuset
60       5.5-1. Cpuset Interface Files
61     5-6. Device
62     5-7. RDMA
63       5-7-1. RDMA Interface Files
64     5-8. Misc
65       5-8-1. perf_event
66     5-N. Non-normative information
67       5-N-1. CPU controller root cgroup process behaviour
68       5-N-2. IO controller root cgroup process behaviour
69   6. Namespace
70     6-1. Basics
71     6-2. The Root and Views
72     6-3. Migration and setns(2)
73     6-4. Interaction with Other Namespaces
74   P. Information on Kernel Programming
75     P-1. Filesystem Support for Writeback
76   D. Deprecated v1 Core Features
77   R. Issues with v1 and Rationales for v2
78     R-1. Multiple Hierarchies
79     R-2. Thread Granularity
80     R-3. Competition Between Inner Nodes and Threads
81     R-4. Other Interface Issues
82     R-5. Controller Issues and Remedies
83       R-5-1. Memory
84
85
86Introduction
87============
88
89Terminology
90-----------
91
92"cgroup" stands for "control group" and is never capitalized.  The
93singular form is used to designate the whole feature and also as a
94qualifier as in "cgroup controllers".  When explicitly referring to
95multiple individual control groups, the plural form "cgroups" is used.
96
97
98What is cgroup?
99---------------
100
101cgroup is a mechanism to organize processes hierarchically and
102distribute system resources along the hierarchy in a controlled and
103configurable manner.
104
105cgroup is largely composed of two parts - the core and controllers.
106cgroup core is primarily responsible for hierarchically organizing
107processes.  A cgroup controller is usually responsible for
108distributing a specific type of system resource along the hierarchy
109although there are utility controllers which serve purposes other than
110resource distribution.
111
112cgroups form a tree structure and every process in the system belongs
113to one and only one cgroup.  All threads of a process belong to the
114same cgroup.  On creation, all processes are put in the cgroup that
115the parent process belongs to at the time.  A process can be migrated
116to another cgroup.  Migration of a process doesn't affect already
117existing descendant processes.
118
119Following certain structural constraints, controllers may be enabled or
120disabled selectively on a cgroup.  All controller behaviors are
121hierarchical - if a controller is enabled on a cgroup, it affects all
122processes which belong to the cgroups consisting the inclusive
123sub-hierarchy of the cgroup.  When a controller is enabled on a nested
124cgroup, it always restricts the resource distribution further.  The
125restrictions set closer to the root in the hierarchy can not be
126overridden from further away.
127
128
129Basic Operations
130================
131
132Mounting
133--------
134
135Unlike v1, cgroup v2 has only single hierarchy.  The cgroup v2
136hierarchy can be mounted with the following mount command::
137
138  # mount -t cgroup2 none $MOUNT_POINT
139
140cgroup2 filesystem has the magic number 0x63677270 ("cgrp").  All
141controllers which support v2 and are not bound to a v1 hierarchy are
142automatically bound to the v2 hierarchy and show up at the root.
143Controllers which are not in active use in the v2 hierarchy can be
144bound to other hierarchies.  This allows mixing v2 hierarchy with the
145legacy v1 multiple hierarchies in a fully backward compatible way.
146
147A controller can be moved across hierarchies only after the controller
148is no longer referenced in its current hierarchy.  Because per-cgroup
149controller states are destroyed asynchronously and controllers may
150have lingering references, a controller may not show up immediately on
151the v2 hierarchy after the final umount of the previous hierarchy.
152Similarly, a controller should be fully disabled to be moved out of
153the unified hierarchy and it may take some time for the disabled
154controller to become available for other hierarchies; furthermore, due
155to inter-controller dependencies, other controllers may need to be
156disabled too.
157
158While useful for development and manual configurations, moving
159controllers dynamically between the v2 and other hierarchies is
160strongly discouraged for production use.  It is recommended to decide
161the hierarchies and controller associations before starting using the
162controllers after system boot.
163
164During transition to v2, system management software might still
165automount the v1 cgroup filesystem and so hijack all controllers
166during boot, before manual intervention is possible. To make testing
167and experimenting easier, the kernel parameter cgroup_no_v1= allows
168disabling controllers in v1 and make them always available in v2.
169
170cgroup v2 currently supports the following mount options.
171
172  nsdelegate
173
174	Consider cgroup namespaces as delegation boundaries.  This
175	option is system wide and can only be set on mount or modified
176	through remount from the init namespace.  The mount option is
177	ignored on non-init namespace mounts.  Please refer to the
178	Delegation section for details.
179
180
181Organizing Processes and Threads
182--------------------------------
183
184Processes
185~~~~~~~~~
186
187Initially, only the root cgroup exists to which all processes belong.
188A child cgroup can be created by creating a sub-directory::
189
190  # mkdir $CGROUP_NAME
191
192A given cgroup may have multiple child cgroups forming a tree
193structure.  Each cgroup has a read-writable interface file
194"cgroup.procs".  When read, it lists the PIDs of all processes which
195belong to the cgroup one-per-line.  The PIDs are not ordered and the
196same PID may show up more than once if the process got moved to
197another cgroup and then back or the PID got recycled while reading.
198
199A process can be migrated into a cgroup by writing its PID to the
200target cgroup's "cgroup.procs" file.  Only one process can be migrated
201on a single write(2) call.  If a process is composed of multiple
202threads, writing the PID of any thread migrates all threads of the
203process.
204
205When a process forks a child process, the new process is born into the
206cgroup that the forking process belongs to at the time of the
207operation.  After exit, a process stays associated with the cgroup
208that it belonged to at the time of exit until it's reaped; however, a
209zombie process does not appear in "cgroup.procs" and thus can't be
210moved to another cgroup.
211
212A cgroup which doesn't have any children or live processes can be
213destroyed by removing the directory.  Note that a cgroup which doesn't
214have any children and is associated only with zombie processes is
215considered empty and can be removed::
216
217  # rmdir $CGROUP_NAME
218
219"/proc/$PID/cgroup" lists a process's cgroup membership.  If legacy
220cgroup is in use in the system, this file may contain multiple lines,
221one for each hierarchy.  The entry for cgroup v2 is always in the
222format "0::$PATH"::
223
224  # cat /proc/842/cgroup
225  ...
226  0::/test-cgroup/test-cgroup-nested
227
228If the process becomes a zombie and the cgroup it was associated with
229is removed subsequently, " (deleted)" is appended to the path::
230
231  # cat /proc/842/cgroup
232  ...
233  0::/test-cgroup/test-cgroup-nested (deleted)
234
235
236Threads
237~~~~~~~
238
239cgroup v2 supports thread granularity for a subset of controllers to
240support use cases requiring hierarchical resource distribution across
241the threads of a group of processes.  By default, all threads of a
242process belong to the same cgroup, which also serves as the resource
243domain to host resource consumptions which are not specific to a
244process or thread.  The thread mode allows threads to be spread across
245a subtree while still maintaining the common resource domain for them.
246
247Controllers which support thread mode are called threaded controllers.
248The ones which don't are called domain controllers.
249
250Marking a cgroup threaded makes it join the resource domain of its
251parent as a threaded cgroup.  The parent may be another threaded
252cgroup whose resource domain is further up in the hierarchy.  The root
253of a threaded subtree, that is, the nearest ancestor which is not
254threaded, is called threaded domain or thread root interchangeably and
255serves as the resource domain for the entire subtree.
256
257Inside a threaded subtree, threads of a process can be put in
258different cgroups and are not subject to the no internal process
259constraint - threaded controllers can be enabled on non-leaf cgroups
260whether they have threads in them or not.
261
262As the threaded domain cgroup hosts all the domain resource
263consumptions of the subtree, it is considered to have internal
264resource consumptions whether there are processes in it or not and
265can't have populated child cgroups which aren't threaded.  Because the
266root cgroup is not subject to no internal process constraint, it can
267serve both as a threaded domain and a parent to domain cgroups.
268
269The current operation mode or type of the cgroup is shown in the
270"cgroup.type" file which indicates whether the cgroup is a normal
271domain, a domain which is serving as the domain of a threaded subtree,
272or a threaded cgroup.
273
274On creation, a cgroup is always a domain cgroup and can be made
275threaded by writing "threaded" to the "cgroup.type" file.  The
276operation is single direction::
277
278  # echo threaded > cgroup.type
279
280Once threaded, the cgroup can't be made a domain again.  To enable the
281thread mode, the following conditions must be met.
282
283- As the cgroup will join the parent's resource domain.  The parent
284  must either be a valid (threaded) domain or a threaded cgroup.
285
286- When the parent is an unthreaded domain, it must not have any domain
287  controllers enabled or populated domain children.  The root is
288  exempt from this requirement.
289
290Topology-wise, a cgroup can be in an invalid state.  Please consider
291the following topology::
292
293  A (threaded domain) - B (threaded) - C (domain, just created)
294
295C is created as a domain but isn't connected to a parent which can
296host child domains.  C can't be used until it is turned into a
297threaded cgroup.  "cgroup.type" file will report "domain (invalid)" in
298these cases.  Operations which fail due to invalid topology use
299EOPNOTSUPP as the errno.
300
301A domain cgroup is turned into a threaded domain when one of its child
302cgroup becomes threaded or threaded controllers are enabled in the
303"cgroup.subtree_control" file while there are processes in the cgroup.
304A threaded domain reverts to a normal domain when the conditions
305clear.
306
307When read, "cgroup.threads" contains the list of the thread IDs of all
308threads in the cgroup.  Except that the operations are per-thread
309instead of per-process, "cgroup.threads" has the same format and
310behaves the same way as "cgroup.procs".  While "cgroup.threads" can be
311written to in any cgroup, as it can only move threads inside the same
312threaded domain, its operations are confined inside each threaded
313subtree.
314
315The threaded domain cgroup serves as the resource domain for the whole
316subtree, and, while the threads can be scattered across the subtree,
317all the processes are considered to be in the threaded domain cgroup.
318"cgroup.procs" in a threaded domain cgroup contains the PIDs of all
319processes in the subtree and is not readable in the subtree proper.
320However, "cgroup.procs" can be written to from anywhere in the subtree
321to migrate all threads of the matching process to the cgroup.
322
323Only threaded controllers can be enabled in a threaded subtree.  When
324a threaded controller is enabled inside a threaded subtree, it only
325accounts for and controls resource consumptions associated with the
326threads in the cgroup and its descendants.  All consumptions which
327aren't tied to a specific thread belong to the threaded domain cgroup.
328
329Because a threaded subtree is exempt from no internal process
330constraint, a threaded controller must be able to handle competition
331between threads in a non-leaf cgroup and its child cgroups.  Each
332threaded controller defines how such competitions are handled.
333
334
335[Un]populated Notification
336--------------------------
337
338Each non-root cgroup has a "cgroup.events" file which contains
339"populated" field indicating whether the cgroup's sub-hierarchy has
340live processes in it.  Its value is 0 if there is no live process in
341the cgroup and its descendants; otherwise, 1.  poll and [id]notify
342events are triggered when the value changes.  This can be used, for
343example, to start a clean-up operation after all processes of a given
344sub-hierarchy have exited.  The populated state updates and
345notifications are recursive.  Consider the following sub-hierarchy
346where the numbers in the parentheses represent the numbers of processes
347in each cgroup::
348
349  A(4) - B(0) - C(1)
350              \ D(0)
351
352A, B and C's "populated" fields would be 1 while D's 0.  After the one
353process in C exits, B and C's "populated" fields would flip to "0" and
354file modified events will be generated on the "cgroup.events" files of
355both cgroups.
356
357
358Controlling Controllers
359-----------------------
360
361Enabling and Disabling
362~~~~~~~~~~~~~~~~~~~~~~
363
364Each cgroup has a "cgroup.controllers" file which lists all
365controllers available for the cgroup to enable::
366
367  # cat cgroup.controllers
368  cpu io memory
369
370No controller is enabled by default.  Controllers can be enabled and
371disabled by writing to the "cgroup.subtree_control" file::
372
373  # echo "+cpu +memory -io" > cgroup.subtree_control
374
375Only controllers which are listed in "cgroup.controllers" can be
376enabled.  When multiple operations are specified as above, either they
377all succeed or fail.  If multiple operations on the same controller
378are specified, the last one is effective.
379
380Enabling a controller in a cgroup indicates that the distribution of
381the target resource across its immediate children will be controlled.
382Consider the following sub-hierarchy.  The enabled controllers are
383listed in parentheses::
384
385  A(cpu,memory) - B(memory) - C()
386                            \ D()
387
388As A has "cpu" and "memory" enabled, A will control the distribution
389of CPU cycles and memory to its children, in this case, B.  As B has
390"memory" enabled but not "CPU", C and D will compete freely on CPU
391cycles but their division of memory available to B will be controlled.
392
393As a controller regulates the distribution of the target resource to
394the cgroup's children, enabling it creates the controller's interface
395files in the child cgroups.  In the above example, enabling "cpu" on B
396would create the "cpu." prefixed controller interface files in C and
397D.  Likewise, disabling "memory" from B would remove the "memory."
398prefixed controller interface files from C and D.  This means that the
399controller interface files - anything which doesn't start with
400"cgroup." are owned by the parent rather than the cgroup itself.
401
402
403Top-down Constraint
404~~~~~~~~~~~~~~~~~~~
405
406Resources are distributed top-down and a cgroup can further distribute
407a resource only if the resource has been distributed to it from the
408parent.  This means that all non-root "cgroup.subtree_control" files
409can only contain controllers which are enabled in the parent's
410"cgroup.subtree_control" file.  A controller can be enabled only if
411the parent has the controller enabled and a controller can't be
412disabled if one or more children have it enabled.
413
414
415No Internal Process Constraint
416~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
417
418Non-root cgroups can distribute domain resources to their children
419only when they don't have any processes of their own.  In other words,
420only domain cgroups which don't contain any processes can have domain
421controllers enabled in their "cgroup.subtree_control" files.
422
423This guarantees that, when a domain controller is looking at the part
424of the hierarchy which has it enabled, processes are always only on
425the leaves.  This rules out situations where child cgroups compete
426against internal processes of the parent.
427
428The root cgroup is exempt from this restriction.  Root contains
429processes and anonymous resource consumption which can't be associated
430with any other cgroups and requires special treatment from most
431controllers.  How resource consumption in the root cgroup is governed
432is up to each controller (for more information on this topic please
433refer to the Non-normative information section in the Controllers
434chapter).
435
436Note that the restriction doesn't get in the way if there is no
437enabled controller in the cgroup's "cgroup.subtree_control".  This is
438important as otherwise it wouldn't be possible to create children of a
439populated cgroup.  To control resource distribution of a cgroup, the
440cgroup must create children and transfer all its processes to the
441children before enabling controllers in its "cgroup.subtree_control"
442file.
443
444
445Delegation
446----------
447
448Model of Delegation
449~~~~~~~~~~~~~~~~~~~
450
451A cgroup can be delegated in two ways.  First, to a less privileged
452user by granting write access of the directory and its "cgroup.procs",
453"cgroup.threads" and "cgroup.subtree_control" files to the user.
454Second, if the "nsdelegate" mount option is set, automatically to a
455cgroup namespace on namespace creation.
456
457Because the resource control interface files in a given directory
458control the distribution of the parent's resources, the delegatee
459shouldn't be allowed to write to them.  For the first method, this is
460achieved by not granting access to these files.  For the second, the
461kernel rejects writes to all files other than "cgroup.procs" and
462"cgroup.subtree_control" on a namespace root from inside the
463namespace.
464
465The end results are equivalent for both delegation types.  Once
466delegated, the user can build sub-hierarchy under the directory,
467organize processes inside it as it sees fit and further distribute the
468resources it received from the parent.  The limits and other settings
469of all resource controllers are hierarchical and regardless of what
470happens in the delegated sub-hierarchy, nothing can escape the
471resource restrictions imposed by the parent.
472
473Currently, cgroup doesn't impose any restrictions on the number of
474cgroups in or nesting depth of a delegated sub-hierarchy; however,
475this may be limited explicitly in the future.
476
477
478Delegation Containment
479~~~~~~~~~~~~~~~~~~~~~~
480
481A delegated sub-hierarchy is contained in the sense that processes
482can't be moved into or out of the sub-hierarchy by the delegatee.
483
484For delegations to a less privileged user, this is achieved by
485requiring the following conditions for a process with a non-root euid
486to migrate a target process into a cgroup by writing its PID to the
487"cgroup.procs" file.
488
489- The writer must have write access to the "cgroup.procs" file.
490
491- The writer must have write access to the "cgroup.procs" file of the
492  common ancestor of the source and destination cgroups.
493
494The above two constraints ensure that while a delegatee may migrate
495processes around freely in the delegated sub-hierarchy it can't pull
496in from or push out to outside the sub-hierarchy.
497
498For an example, let's assume cgroups C0 and C1 have been delegated to
499user U0 who created C00, C01 under C0 and C10 under C1 as follows and
500all processes under C0 and C1 belong to U0::
501
502  ~~~~~~~~~~~~~ - C0 - C00
503  ~ cgroup    ~      \ C01
504  ~ hierarchy ~
505  ~~~~~~~~~~~~~ - C1 - C10
506
507Let's also say U0 wants to write the PID of a process which is
508currently in C10 into "C00/cgroup.procs".  U0 has write access to the
509file; however, the common ancestor of the source cgroup C10 and the
510destination cgroup C00 is above the points of delegation and U0 would
511not have write access to its "cgroup.procs" files and thus the write
512will be denied with -EACCES.
513
514For delegations to namespaces, containment is achieved by requiring
515that both the source and destination cgroups are reachable from the
516namespace of the process which is attempting the migration.  If either
517is not reachable, the migration is rejected with -ENOENT.
518
519
520Guidelines
521----------
522
523Organize Once and Control
524~~~~~~~~~~~~~~~~~~~~~~~~~
525
526Migrating a process across cgroups is a relatively expensive operation
527and stateful resources such as memory are not moved together with the
528process.  This is an explicit design decision as there often exist
529inherent trade-offs between migration and various hot paths in terms
530of synchronization cost.
531
532As such, migrating processes across cgroups frequently as a means to
533apply different resource restrictions is discouraged.  A workload
534should be assigned to a cgroup according to the system's logical and
535resource structure once on start-up.  Dynamic adjustments to resource
536distribution can be made by changing controller configuration through
537the interface files.
538
539
540Avoid Name Collisions
541~~~~~~~~~~~~~~~~~~~~~
542
543Interface files for a cgroup and its children cgroups occupy the same
544directory and it is possible to create children cgroups which collide
545with interface files.
546
547All cgroup core interface files are prefixed with "cgroup." and each
548controller's interface files are prefixed with the controller name and
549a dot.  A controller's name is composed of lower case alphabets and
550'_'s but never begins with an '_' so it can be used as the prefix
551character for collision avoidance.  Also, interface file names won't
552start or end with terms which are often used in categorizing workloads
553such as job, service, slice, unit or workload.
554
555cgroup doesn't do anything to prevent name collisions and it's the
556user's responsibility to avoid them.
557
558
559Resource Distribution Models
560============================
561
562cgroup controllers implement several resource distribution schemes
563depending on the resource type and expected use cases.  This section
564describes major schemes in use along with their expected behaviors.
565
566
567Weights
568-------
569
570A parent's resource is distributed by adding up the weights of all
571active children and giving each the fraction matching the ratio of its
572weight against the sum.  As only children which can make use of the
573resource at the moment participate in the distribution, this is
574work-conserving.  Due to the dynamic nature, this model is usually
575used for stateless resources.
576
577All weights are in the range [1, 10000] with the default at 100.  This
578allows symmetric multiplicative biases in both directions at fine
579enough granularity while staying in the intuitive range.
580
581As long as the weight is in range, all configuration combinations are
582valid and there is no reason to reject configuration changes or
583process migrations.
584
585"cpu.weight" proportionally distributes CPU cycles to active children
586and is an example of this type.
587
588
589Limits
590------
591
592A child can only consume upto the configured amount of the resource.
593Limits can be over-committed - the sum of the limits of children can
594exceed the amount of resource available to the parent.
595
596Limits are in the range [0, max] and defaults to "max", which is noop.
597
598As limits can be over-committed, all configuration combinations are
599valid and there is no reason to reject configuration changes or
600process migrations.
601
602"io.max" limits the maximum BPS and/or IOPS that a cgroup can consume
603on an IO device and is an example of this type.
604
605
606Protections
607-----------
608
609A cgroup is protected to be allocated upto the configured amount of
610the resource if the usages of all its ancestors are under their
611protected levels.  Protections can be hard guarantees or best effort
612soft boundaries.  Protections can also be over-committed in which case
613only upto the amount available to the parent is protected among
614children.
615
616Protections are in the range [0, max] and defaults to 0, which is
617noop.
618
619As protections can be over-committed, all configuration combinations
620are valid and there is no reason to reject configuration changes or
621process migrations.
622
623"memory.low" implements best-effort memory protection and is an
624example of this type.
625
626
627Allocations
628-----------
629
630A cgroup is exclusively allocated a certain amount of a finite
631resource.  Allocations can't be over-committed - the sum of the
632allocations of children can not exceed the amount of resource
633available to the parent.
634
635Allocations are in the range [0, max] and defaults to 0, which is no
636resource.
637
638As allocations can't be over-committed, some configuration
639combinations are invalid and should be rejected.  Also, if the
640resource is mandatory for execution of processes, process migrations
641may be rejected.
642
643"cpu.rt.max" hard-allocates realtime slices and is an example of this
644type.
645
646
647Interface Files
648===============
649
650Format
651------
652
653All interface files should be in one of the following formats whenever
654possible::
655
656  New-line separated values
657  (when only one value can be written at once)
658
659	VAL0\n
660	VAL1\n
661	...
662
663  Space separated values
664  (when read-only or multiple values can be written at once)
665
666	VAL0 VAL1 ...\n
667
668  Flat keyed
669
670	KEY0 VAL0\n
671	KEY1 VAL1\n
672	...
673
674  Nested keyed
675
676	KEY0 SUB_KEY0=VAL00 SUB_KEY1=VAL01...
677	KEY1 SUB_KEY0=VAL10 SUB_KEY1=VAL11...
678	...
679
680For a writable file, the format for writing should generally match
681reading; however, controllers may allow omitting later fields or
682implement restricted shortcuts for most common use cases.
683
684For both flat and nested keyed files, only the values for a single key
685can be written at a time.  For nested keyed files, the sub key pairs
686may be specified in any order and not all pairs have to be specified.
687
688
689Conventions
690-----------
691
692- Settings for a single feature should be contained in a single file.
693
694- The root cgroup should be exempt from resource control and thus
695  shouldn't have resource control interface files.  Also,
696  informational files on the root cgroup which end up showing global
697  information available elsewhere shouldn't exist.
698
699- If a controller implements weight based resource distribution, its
700  interface file should be named "weight" and have the range [1,
701  10000] with 100 as the default.  The values are chosen to allow
702  enough and symmetric bias in both directions while keeping it
703  intuitive (the default is 100%).
704
705- If a controller implements an absolute resource guarantee and/or
706  limit, the interface files should be named "min" and "max"
707  respectively.  If a controller implements best effort resource
708  guarantee and/or limit, the interface files should be named "low"
709  and "high" respectively.
710
711  In the above four control files, the special token "max" should be
712  used to represent upward infinity for both reading and writing.
713
714- If a setting has a configurable default value and keyed specific
715  overrides, the default entry should be keyed with "default" and
716  appear as the first entry in the file.
717
718  The default value can be updated by writing either "default $VAL" or
719  "$VAL".
720
721  When writing to update a specific override, "default" can be used as
722  the value to indicate removal of the override.  Override entries
723  with "default" as the value must not appear when read.
724
725  For example, a setting which is keyed by major:minor device numbers
726  with integer values may look like the following::
727
728    # cat cgroup-example-interface-file
729    default 150
730    8:0 300
731
732  The default value can be updated by::
733
734    # echo 125 > cgroup-example-interface-file
735
736  or::
737
738    # echo "default 125" > cgroup-example-interface-file
739
740  An override can be set by::
741
742    # echo "8:16 170" > cgroup-example-interface-file
743
744  and cleared by::
745
746    # echo "8:0 default" > cgroup-example-interface-file
747    # cat cgroup-example-interface-file
748    default 125
749    8:16 170
750
751- For events which are not very high frequency, an interface file
752  "events" should be created which lists event key value pairs.
753  Whenever a notifiable event happens, file modified event should be
754  generated on the file.
755
756
757Core Interface Files
758--------------------
759
760All cgroup core files are prefixed with "cgroup."
761
762  cgroup.type
763
764	A read-write single value file which exists on non-root
765	cgroups.
766
767	When read, it indicates the current type of the cgroup, which
768	can be one of the following values.
769
770	- "domain" : A normal valid domain cgroup.
771
772	- "domain threaded" : A threaded domain cgroup which is
773          serving as the root of a threaded subtree.
774
775	- "domain invalid" : A cgroup which is in an invalid state.
776	  It can't be populated or have controllers enabled.  It may
777	  be allowed to become a threaded cgroup.
778
779	- "threaded" : A threaded cgroup which is a member of a
780          threaded subtree.
781
782	A cgroup can be turned into a threaded cgroup by writing
783	"threaded" to this file.
784
785  cgroup.procs
786	A read-write new-line separated values file which exists on
787	all cgroups.
788
789	When read, it lists the PIDs of all processes which belong to
790	the cgroup one-per-line.  The PIDs are not ordered and the
791	same PID may show up more than once if the process got moved
792	to another cgroup and then back or the PID got recycled while
793	reading.
794
795	A PID can be written to migrate the process associated with
796	the PID to the cgroup.  The writer should match all of the
797	following conditions.
798
799	- It must have write access to the "cgroup.procs" file.
800
801	- It must have write access to the "cgroup.procs" file of the
802	  common ancestor of the source and destination cgroups.
803
804	When delegating a sub-hierarchy, write access to this file
805	should be granted along with the containing directory.
806
807	In a threaded cgroup, reading this file fails with EOPNOTSUPP
808	as all the processes belong to the thread root.  Writing is
809	supported and moves every thread of the process to the cgroup.
810
811  cgroup.threads
812	A read-write new-line separated values file which exists on
813	all cgroups.
814
815	When read, it lists the TIDs of all threads which belong to
816	the cgroup one-per-line.  The TIDs are not ordered and the
817	same TID may show up more than once if the thread got moved to
818	another cgroup and then back or the TID got recycled while
819	reading.
820
821	A TID can be written to migrate the thread associated with the
822	TID to the cgroup.  The writer should match all of the
823	following conditions.
824
825	- It must have write access to the "cgroup.threads" file.
826
827	- The cgroup that the thread is currently in must be in the
828          same resource domain as the destination cgroup.
829
830	- It must have write access to the "cgroup.procs" file of the
831	  common ancestor of the source and destination cgroups.
832
833	When delegating a sub-hierarchy, write access to this file
834	should be granted along with the containing directory.
835
836  cgroup.controllers
837	A read-only space separated values file which exists on all
838	cgroups.
839
840	It shows space separated list of all controllers available to
841	the cgroup.  The controllers are not ordered.
842
843  cgroup.subtree_control
844	A read-write space separated values file which exists on all
845	cgroups.  Starts out empty.
846
847	When read, it shows space separated list of the controllers
848	which are enabled to control resource distribution from the
849	cgroup to its children.
850
851	Space separated list of controllers prefixed with '+' or '-'
852	can be written to enable or disable controllers.  A controller
853	name prefixed with '+' enables the controller and '-'
854	disables.  If a controller appears more than once on the list,
855	the last one is effective.  When multiple enable and disable
856	operations are specified, either all succeed or all fail.
857
858  cgroup.events
859	A read-only flat-keyed file which exists on non-root cgroups.
860	The following entries are defined.  Unless specified
861	otherwise, a value change in this file generates a file
862	modified event.
863
864	  populated
865		1 if the cgroup or its descendants contains any live
866		processes; otherwise, 0.
867
868  cgroup.max.descendants
869	A read-write single value files.  The default is "max".
870
871	Maximum allowed number of descent cgroups.
872	If the actual number of descendants is equal or larger,
873	an attempt to create a new cgroup in the hierarchy will fail.
874
875  cgroup.max.depth
876	A read-write single value files.  The default is "max".
877
878	Maximum allowed descent depth below the current cgroup.
879	If the actual descent depth is equal or larger,
880	an attempt to create a new child cgroup will fail.
881
882  cgroup.stat
883	A read-only flat-keyed file with the following entries:
884
885	  nr_descendants
886		Total number of visible descendant cgroups.
887
888	  nr_dying_descendants
889		Total number of dying descendant cgroups. A cgroup becomes
890		dying after being deleted by a user. The cgroup will remain
891		in dying state for some time undefined time (which can depend
892		on system load) before being completely destroyed.
893
894		A process can't enter a dying cgroup under any circumstances,
895		a dying cgroup can't revive.
896
897		A dying cgroup can consume system resources not exceeding
898		limits, which were active at the moment of cgroup deletion.
899
900
901Controllers
902===========
903
904CPU
905---
906
907The "cpu" controllers regulates distribution of CPU cycles.  This
908controller implements weight and absolute bandwidth limit models for
909normal scheduling policy and absolute bandwidth allocation model for
910realtime scheduling policy.
911
912WARNING: cgroup2 doesn't yet support control of realtime processes and
913the cpu controller can only be enabled when all RT processes are in
914the root cgroup.  Be aware that system management software may already
915have placed RT processes into nonroot cgroups during the system boot
916process, and these processes may need to be moved to the root cgroup
917before the cpu controller can be enabled.
918
919
920CPU Interface Files
921~~~~~~~~~~~~~~~~~~~
922
923All time durations are in microseconds.
924
925  cpu.stat
926	A read-only flat-keyed file which exists on non-root cgroups.
927	This file exists whether the controller is enabled or not.
928
929	It always reports the following three stats:
930
931	- usage_usec
932	- user_usec
933	- system_usec
934
935	and the following three when the controller is enabled:
936
937	- nr_periods
938	- nr_throttled
939	- throttled_usec
940
941  cpu.weight
942	A read-write single value file which exists on non-root
943	cgroups.  The default is "100".
944
945	The weight in the range [1, 10000].
946
947  cpu.weight.nice
948	A read-write single value file which exists on non-root
949	cgroups.  The default is "0".
950
951	The nice value is in the range [-20, 19].
952
953	This interface file is an alternative interface for
954	"cpu.weight" and allows reading and setting weight using the
955	same values used by nice(2).  Because the range is smaller and
956	granularity is coarser for the nice values, the read value is
957	the closest approximation of the current weight.
958
959  cpu.max
960	A read-write two value file which exists on non-root cgroups.
961	The default is "max 100000".
962
963	The maximum bandwidth limit.  It's in the following format::
964
965	  $MAX $PERIOD
966
967	which indicates that the group may consume upto $MAX in each
968	$PERIOD duration.  "max" for $MAX indicates no limit.  If only
969	one number is written, $MAX is updated.
970
971  cpu.pressure
972	A read-only nested-key file which exists on non-root cgroups.
973
974	Shows pressure stall information for CPU. See
975	Documentation/accounting/psi.txt for details.
976
977
978Memory
979------
980
981The "memory" controller regulates distribution of memory.  Memory is
982stateful and implements both limit and protection models.  Due to the
983intertwining between memory usage and reclaim pressure and the
984stateful nature of memory, the distribution model is relatively
985complex.
986
987While not completely water-tight, all major memory usages by a given
988cgroup are tracked so that the total memory consumption can be
989accounted and controlled to a reasonable extent.  Currently, the
990following types of memory usages are tracked.
991
992- Userland memory - page cache and anonymous memory.
993
994- Kernel data structures such as dentries and inodes.
995
996- TCP socket buffers.
997
998The above list may expand in the future for better coverage.
999
1000
1001Memory Interface Files
1002~~~~~~~~~~~~~~~~~~~~~~
1003
1004All memory amounts are in bytes.  If a value which is not aligned to
1005PAGE_SIZE is written, the value may be rounded up to the closest
1006PAGE_SIZE multiple when read back.
1007
1008  memory.current
1009	A read-only single value file which exists on non-root
1010	cgroups.
1011
1012	The total amount of memory currently being used by the cgroup
1013	and its descendants.
1014
1015  memory.min
1016	A read-write single value file which exists on non-root
1017	cgroups.  The default is "0".
1018
1019	Hard memory protection.  If the memory usage of a cgroup
1020	is within its effective min boundary, the cgroup's memory
1021	won't be reclaimed under any conditions. If there is no
1022	unprotected reclaimable memory available, OOM killer
1023	is invoked.
1024
1025       Effective min boundary is limited by memory.min values of
1026	all ancestor cgroups. If there is memory.min overcommitment
1027	(child cgroup or cgroups are requiring more protected memory
1028	than parent will allow), then each child cgroup will get
1029	the part of parent's protection proportional to its
1030	actual memory usage below memory.min.
1031
1032	Putting more memory than generally available under this
1033	protection is discouraged and may lead to constant OOMs.
1034
1035	If a memory cgroup is not populated with processes,
1036	its memory.min is ignored.
1037
1038  memory.low
1039	A read-write single value file which exists on non-root
1040	cgroups.  The default is "0".
1041
1042	Best-effort memory protection.  If the memory usage of a
1043	cgroup is within its effective low boundary, the cgroup's
1044	memory won't be reclaimed unless memory can be reclaimed
1045	from unprotected cgroups.
1046
1047	Effective low boundary is limited by memory.low values of
1048	all ancestor cgroups. If there is memory.low overcommitment
1049	(child cgroup or cgroups are requiring more protected memory
1050	than parent will allow), then each child cgroup will get
1051	the part of parent's protection proportional to its
1052	actual memory usage below memory.low.
1053
1054	Putting more memory than generally available under this
1055	protection is discouraged.
1056
1057  memory.high
1058	A read-write single value file which exists on non-root
1059	cgroups.  The default is "max".
1060
1061	Memory usage throttle limit.  This is the main mechanism to
1062	control memory usage of a cgroup.  If a cgroup's usage goes
1063	over the high boundary, the processes of the cgroup are
1064	throttled and put under heavy reclaim pressure.
1065
1066	Going over the high limit never invokes the OOM killer and
1067	under extreme conditions the limit may be breached.
1068
1069  memory.max
1070	A read-write single value file which exists on non-root
1071	cgroups.  The default is "max".
1072
1073	Memory usage hard limit.  This is the final protection
1074	mechanism.  If a cgroup's memory usage reaches this limit and
1075	can't be reduced, the OOM killer is invoked in the cgroup.
1076	Under certain circumstances, the usage may go over the limit
1077	temporarily.
1078
1079	This is the ultimate protection mechanism.  As long as the
1080	high limit is used and monitored properly, this limit's
1081	utility is limited to providing the final safety net.
1082
1083  memory.oom.group
1084	A read-write single value file which exists on non-root
1085	cgroups.  The default value is "0".
1086
1087	Determines whether the cgroup should be treated as
1088	an indivisible workload by the OOM killer. If set,
1089	all tasks belonging to the cgroup or to its descendants
1090	(if the memory cgroup is not a leaf cgroup) are killed
1091	together or not at all. This can be used to avoid
1092	partial kills to guarantee workload integrity.
1093
1094	Tasks with the OOM protection (oom_score_adj set to -1000)
1095	are treated as an exception and are never killed.
1096
1097	If the OOM killer is invoked in a cgroup, it's not going
1098	to kill any tasks outside of this cgroup, regardless
1099	memory.oom.group values of ancestor cgroups.
1100
1101  memory.events
1102	A read-only flat-keyed file which exists on non-root cgroups.
1103	The following entries are defined.  Unless specified
1104	otherwise, a value change in this file generates a file
1105	modified event.
1106
1107	  low
1108		The number of times the cgroup is reclaimed due to
1109		high memory pressure even though its usage is under
1110		the low boundary.  This usually indicates that the low
1111		boundary is over-committed.
1112
1113	  high
1114		The number of times processes of the cgroup are
1115		throttled and routed to perform direct memory reclaim
1116		because the high memory boundary was exceeded.  For a
1117		cgroup whose memory usage is capped by the high limit
1118		rather than global memory pressure, this event's
1119		occurrences are expected.
1120
1121	  max
1122		The number of times the cgroup's memory usage was
1123		about to go over the max boundary.  If direct reclaim
1124		fails to bring it down, the cgroup goes to OOM state.
1125
1126	  oom
1127		The number of time the cgroup's memory usage was
1128		reached the limit and allocation was about to fail.
1129
1130		Depending on context result could be invocation of OOM
1131		killer and retrying allocation or failing allocation.
1132
1133		Failed allocation in its turn could be returned into
1134		userspace as -ENOMEM or silently ignored in cases like
1135		disk readahead.  For now OOM in memory cgroup kills
1136		tasks iff shortage has happened inside page fault.
1137
1138		This event is not raised if the OOM killer is not
1139		considered as an option, e.g. for failed high-order
1140		allocations.
1141
1142	  oom_kill
1143		The number of processes belonging to this cgroup
1144		killed by any kind of OOM killer.
1145
1146  memory.stat
1147	A read-only flat-keyed file which exists on non-root cgroups.
1148
1149	This breaks down the cgroup's memory footprint into different
1150	types of memory, type-specific details, and other information
1151	on the state and past events of the memory management system.
1152
1153	All memory amounts are in bytes.
1154
1155	The entries are ordered to be human readable, and new entries
1156	can show up in the middle. Don't rely on items remaining in a
1157	fixed position; use the keys to look up specific values!
1158
1159	  anon
1160		Amount of memory used in anonymous mappings such as
1161		brk(), sbrk(), and mmap(MAP_ANONYMOUS)
1162
1163	  file
1164		Amount of memory used to cache filesystem data,
1165		including tmpfs and shared memory.
1166
1167	  kernel_stack
1168		Amount of memory allocated to kernel stacks.
1169
1170	  slab
1171		Amount of memory used for storing in-kernel data
1172		structures.
1173
1174	  sock
1175		Amount of memory used in network transmission buffers
1176
1177	  shmem
1178		Amount of cached filesystem data that is swap-backed,
1179		such as tmpfs, shm segments, shared anonymous mmap()s
1180
1181	  file_mapped
1182		Amount of cached filesystem data mapped with mmap()
1183
1184	  file_dirty
1185		Amount of cached filesystem data that was modified but
1186		not yet written back to disk
1187
1188	  file_writeback
1189		Amount of cached filesystem data that was modified and
1190		is currently being written back to disk
1191
1192	  anon_thp
1193		Amount of memory used in anonymous mappings backed by
1194		transparent hugepages
1195
1196	  inactive_anon, active_anon, inactive_file, active_file, unevictable
1197		Amount of memory, swap-backed and filesystem-backed,
1198		on the internal memory management lists used by the
1199		page reclaim algorithm
1200
1201	  slab_reclaimable
1202		Part of "slab" that might be reclaimed, such as
1203		dentries and inodes.
1204
1205	  slab_unreclaimable
1206		Part of "slab" that cannot be reclaimed on memory
1207		pressure.
1208
1209	  pgfault
1210		Total number of page faults incurred
1211
1212	  pgmajfault
1213		Number of major page faults incurred
1214
1215	  workingset_refault
1216
1217		Number of refaults of previously evicted pages
1218
1219	  workingset_activate
1220
1221		Number of refaulted pages that were immediately activated
1222
1223	  workingset_nodereclaim
1224
1225		Number of times a shadow node has been reclaimed
1226
1227	  pgrefill
1228
1229		Amount of scanned pages (in an active LRU list)
1230
1231	  pgscan
1232
1233		Amount of scanned pages (in an inactive LRU list)
1234
1235	  pgsteal
1236
1237		Amount of reclaimed pages
1238
1239	  pgactivate
1240
1241		Amount of pages moved to the active LRU list
1242
1243	  pgdeactivate
1244
1245		Amount of pages moved to the inactive LRU lis
1246
1247	  pglazyfree
1248
1249		Amount of pages postponed to be freed under memory pressure
1250
1251	  pglazyfreed
1252
1253		Amount of reclaimed lazyfree pages
1254
1255	  thp_fault_alloc
1256
1257		Number of transparent hugepages which were allocated to satisfy
1258		a page fault, including COW faults. This counter is not present
1259		when CONFIG_TRANSPARENT_HUGEPAGE is not set.
1260
1261	  thp_collapse_alloc
1262
1263		Number of transparent hugepages which were allocated to allow
1264		collapsing an existing range of pages. This counter is not
1265		present when CONFIG_TRANSPARENT_HUGEPAGE is not set.
1266
1267  memory.swap.current
1268	A read-only single value file which exists on non-root
1269	cgroups.
1270
1271	The total amount of swap currently being used by the cgroup
1272	and its descendants.
1273
1274  memory.swap.max
1275	A read-write single value file which exists on non-root
1276	cgroups.  The default is "max".
1277
1278	Swap usage hard limit.  If a cgroup's swap usage reaches this
1279	limit, anonymous memory of the cgroup will not be swapped out.
1280
1281  memory.swap.events
1282	A read-only flat-keyed file which exists on non-root cgroups.
1283	The following entries are defined.  Unless specified
1284	otherwise, a value change in this file generates a file
1285	modified event.
1286
1287	  max
1288		The number of times the cgroup's swap usage was about
1289		to go over the max boundary and swap allocation
1290		failed.
1291
1292	  fail
1293		The number of times swap allocation failed either
1294		because of running out of swap system-wide or max
1295		limit.
1296
1297	When reduced under the current usage, the existing swap
1298	entries are reclaimed gradually and the swap usage may stay
1299	higher than the limit for an extended period of time.  This
1300	reduces the impact on the workload and memory management.
1301
1302  memory.pressure
1303	A read-only nested-key file which exists on non-root cgroups.
1304
1305	Shows pressure stall information for memory. See
1306	Documentation/accounting/psi.txt for details.
1307
1308
1309Usage Guidelines
1310~~~~~~~~~~~~~~~~
1311
1312"memory.high" is the main mechanism to control memory usage.
1313Over-committing on high limit (sum of high limits > available memory)
1314and letting global memory pressure to distribute memory according to
1315usage is a viable strategy.
1316
1317Because breach of the high limit doesn't trigger the OOM killer but
1318throttles the offending cgroup, a management agent has ample
1319opportunities to monitor and take appropriate actions such as granting
1320more memory or terminating the workload.
1321
1322Determining whether a cgroup has enough memory is not trivial as
1323memory usage doesn't indicate whether the workload can benefit from
1324more memory.  For example, a workload which writes data received from
1325network to a file can use all available memory but can also operate as
1326performant with a small amount of memory.  A measure of memory
1327pressure - how much the workload is being impacted due to lack of
1328memory - is necessary to determine whether a workload needs more
1329memory; unfortunately, memory pressure monitoring mechanism isn't
1330implemented yet.
1331
1332
1333Memory Ownership
1334~~~~~~~~~~~~~~~~
1335
1336A memory area is charged to the cgroup which instantiated it and stays
1337charged to the cgroup until the area is released.  Migrating a process
1338to a different cgroup doesn't move the memory usages that it
1339instantiated while in the previous cgroup to the new cgroup.
1340
1341A memory area may be used by processes belonging to different cgroups.
1342To which cgroup the area will be charged is in-deterministic; however,
1343over time, the memory area is likely to end up in a cgroup which has
1344enough memory allowance to avoid high reclaim pressure.
1345
1346If a cgroup sweeps a considerable amount of memory which is expected
1347to be accessed repeatedly by other cgroups, it may make sense to use
1348POSIX_FADV_DONTNEED to relinquish the ownership of memory areas
1349belonging to the affected files to ensure correct memory ownership.
1350
1351
1352IO
1353--
1354
1355The "io" controller regulates the distribution of IO resources.  This
1356controller implements both weight based and absolute bandwidth or IOPS
1357limit distribution; however, weight based distribution is available
1358only if cfq-iosched is in use and neither scheme is available for
1359blk-mq devices.
1360
1361
1362IO Interface Files
1363~~~~~~~~~~~~~~~~~~
1364
1365  io.stat
1366	A read-only nested-keyed file which exists on non-root
1367	cgroups.
1368
1369	Lines are keyed by $MAJ:$MIN device numbers and not ordered.
1370	The following nested keys are defined.
1371
1372	  ======	=====================
1373	  rbytes	Bytes read
1374	  wbytes	Bytes written
1375	  rios		Number of read IOs
1376	  wios		Number of write IOs
1377	  dbytes	Bytes discarded
1378	  dios		Number of discard IOs
1379	  ======	=====================
1380
1381	An example read output follows:
1382
1383	  8:16 rbytes=1459200 wbytes=314773504 rios=192 wios=353 dbytes=0 dios=0
1384	  8:0 rbytes=90430464 wbytes=299008000 rios=8950 wios=1252 dbytes=50331648 dios=3021
1385
1386  io.weight
1387	A read-write flat-keyed file which exists on non-root cgroups.
1388	The default is "default 100".
1389
1390	The first line is the default weight applied to devices
1391	without specific override.  The rest are overrides keyed by
1392	$MAJ:$MIN device numbers and not ordered.  The weights are in
1393	the range [1, 10000] and specifies the relative amount IO time
1394	the cgroup can use in relation to its siblings.
1395
1396	The default weight can be updated by writing either "default
1397	$WEIGHT" or simply "$WEIGHT".  Overrides can be set by writing
1398	"$MAJ:$MIN $WEIGHT" and unset by writing "$MAJ:$MIN default".
1399
1400	An example read output follows::
1401
1402	  default 100
1403	  8:16 200
1404	  8:0 50
1405
1406  io.max
1407	A read-write nested-keyed file which exists on non-root
1408	cgroups.
1409
1410	BPS and IOPS based IO limit.  Lines are keyed by $MAJ:$MIN
1411	device numbers and not ordered.  The following nested keys are
1412	defined.
1413
1414	  =====		==================================
1415	  rbps		Max read bytes per second
1416	  wbps		Max write bytes per second
1417	  riops		Max read IO operations per second
1418	  wiops		Max write IO operations per second
1419	  =====		==================================
1420
1421	When writing, any number of nested key-value pairs can be
1422	specified in any order.  "max" can be specified as the value
1423	to remove a specific limit.  If the same key is specified
1424	multiple times, the outcome is undefined.
1425
1426	BPS and IOPS are measured in each IO direction and IOs are
1427	delayed if limit is reached.  Temporary bursts are allowed.
1428
1429	Setting read limit at 2M BPS and write at 120 IOPS for 8:16::
1430
1431	  echo "8:16 rbps=2097152 wiops=120" > io.max
1432
1433	Reading returns the following::
1434
1435	  8:16 rbps=2097152 wbps=max riops=max wiops=120
1436
1437	Write IOPS limit can be removed by writing the following::
1438
1439	  echo "8:16 wiops=max" > io.max
1440
1441	Reading now returns the following::
1442
1443	  8:16 rbps=2097152 wbps=max riops=max wiops=max
1444
1445  io.pressure
1446	A read-only nested-key file which exists on non-root cgroups.
1447
1448	Shows pressure stall information for IO. See
1449	Documentation/accounting/psi.txt for details.
1450
1451
1452Writeback
1453~~~~~~~~~
1454
1455Page cache is dirtied through buffered writes and shared mmaps and
1456written asynchronously to the backing filesystem by the writeback
1457mechanism.  Writeback sits between the memory and IO domains and
1458regulates the proportion of dirty memory by balancing dirtying and
1459write IOs.
1460
1461The io controller, in conjunction with the memory controller,
1462implements control of page cache writeback IOs.  The memory controller
1463defines the memory domain that dirty memory ratio is calculated and
1464maintained for and the io controller defines the io domain which
1465writes out dirty pages for the memory domain.  Both system-wide and
1466per-cgroup dirty memory states are examined and the more restrictive
1467of the two is enforced.
1468
1469cgroup writeback requires explicit support from the underlying
1470filesystem.  Currently, cgroup writeback is implemented on ext2, ext4
1471and btrfs.  On other filesystems, all writeback IOs are attributed to
1472the root cgroup.
1473
1474There are inherent differences in memory and writeback management
1475which affects how cgroup ownership is tracked.  Memory is tracked per
1476page while writeback per inode.  For the purpose of writeback, an
1477inode is assigned to a cgroup and all IO requests to write dirty pages
1478from the inode are attributed to that cgroup.
1479
1480As cgroup ownership for memory is tracked per page, there can be pages
1481which are associated with different cgroups than the one the inode is
1482associated with.  These are called foreign pages.  The writeback
1483constantly keeps track of foreign pages and, if a particular foreign
1484cgroup becomes the majority over a certain period of time, switches
1485the ownership of the inode to that cgroup.
1486
1487While this model is enough for most use cases where a given inode is
1488mostly dirtied by a single cgroup even when the main writing cgroup
1489changes over time, use cases where multiple cgroups write to a single
1490inode simultaneously are not supported well.  In such circumstances, a
1491significant portion of IOs are likely to be attributed incorrectly.
1492As memory controller assigns page ownership on the first use and
1493doesn't update it until the page is released, even if writeback
1494strictly follows page ownership, multiple cgroups dirtying overlapping
1495areas wouldn't work as expected.  It's recommended to avoid such usage
1496patterns.
1497
1498The sysctl knobs which affect writeback behavior are applied to cgroup
1499writeback as follows.
1500
1501  vm.dirty_background_ratio, vm.dirty_ratio
1502	These ratios apply the same to cgroup writeback with the
1503	amount of available memory capped by limits imposed by the
1504	memory controller and system-wide clean memory.
1505
1506  vm.dirty_background_bytes, vm.dirty_bytes
1507	For cgroup writeback, this is calculated into ratio against
1508	total available memory and applied the same way as
1509	vm.dirty[_background]_ratio.
1510
1511
1512IO Latency
1513~~~~~~~~~~
1514
1515This is a cgroup v2 controller for IO workload protection.  You provide a group
1516with a latency target, and if the average latency exceeds that target the
1517controller will throttle any peers that have a lower latency target than the
1518protected workload.
1519
1520The limits are only applied at the peer level in the hierarchy.  This means that
1521in the diagram below, only groups A, B, and C will influence each other, and
1522groups D and F will influence each other.  Group G will influence nobody::
1523
1524			[root]
1525		/	   |		\
1526		A	   B		C
1527	       /  \        |
1528	      D    F	   G
1529
1530
1531So the ideal way to configure this is to set io.latency in groups A, B, and C.
1532Generally you do not want to set a value lower than the latency your device
1533supports.  Experiment to find the value that works best for your workload.
1534Start at higher than the expected latency for your device and watch the
1535avg_lat value in io.stat for your workload group to get an idea of the
1536latency you see during normal operation.  Use the avg_lat value as a basis for
1537your real setting, setting at 10-15% higher than the value in io.stat.
1538
1539How IO Latency Throttling Works
1540~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1541
1542io.latency is work conserving; so as long as everybody is meeting their latency
1543target the controller doesn't do anything.  Once a group starts missing its
1544target it begins throttling any peer group that has a higher target than itself.
1545This throttling takes 2 forms:
1546
1547- Queue depth throttling.  This is the number of outstanding IO's a group is
1548  allowed to have.  We will clamp down relatively quickly, starting at no limit
1549  and going all the way down to 1 IO at a time.
1550
1551- Artificial delay induction.  There are certain types of IO that cannot be
1552  throttled without possibly adversely affecting higher priority groups.  This
1553  includes swapping and metadata IO.  These types of IO are allowed to occur
1554  normally, however they are "charged" to the originating group.  If the
1555  originating group is being throttled you will see the use_delay and delay
1556  fields in io.stat increase.  The delay value is how many microseconds that are
1557  being added to any process that runs in this group.  Because this number can
1558  grow quite large if there is a lot of swapping or metadata IO occurring we
1559  limit the individual delay events to 1 second at a time.
1560
1561Once the victimized group starts meeting its latency target again it will start
1562unthrottling any peer groups that were throttled previously.  If the victimized
1563group simply stops doing IO the global counter will unthrottle appropriately.
1564
1565IO Latency Interface Files
1566~~~~~~~~~~~~~~~~~~~~~~~~~~
1567
1568  io.latency
1569	This takes a similar format as the other controllers.
1570
1571		"MAJOR:MINOR target=<target time in microseconds"
1572
1573  io.stat
1574	If the controller is enabled you will see extra stats in io.stat in
1575	addition to the normal ones.
1576
1577	  depth
1578		This is the current queue depth for the group.
1579
1580	  avg_lat
1581		This is an exponential moving average with a decay rate of 1/exp
1582		bound by the sampling interval.  The decay rate interval can be
1583		calculated by multiplying the win value in io.stat by the
1584		corresponding number of samples based on the win value.
1585
1586	  win
1587		The sampling window size in milliseconds.  This is the minimum
1588		duration of time between evaluation events.  Windows only elapse
1589		with IO activity.  Idle periods extend the most recent window.
1590
1591PID
1592---
1593
1594The process number controller is used to allow a cgroup to stop any
1595new tasks from being fork()'d or clone()'d after a specified limit is
1596reached.
1597
1598The number of tasks in a cgroup can be exhausted in ways which other
1599controllers cannot prevent, thus warranting its own controller.  For
1600example, a fork bomb is likely to exhaust the number of tasks before
1601hitting memory restrictions.
1602
1603Note that PIDs used in this controller refer to TIDs, process IDs as
1604used by the kernel.
1605
1606
1607PID Interface Files
1608~~~~~~~~~~~~~~~~~~~
1609
1610  pids.max
1611	A read-write single value file which exists on non-root
1612	cgroups.  The default is "max".
1613
1614	Hard limit of number of processes.
1615
1616  pids.current
1617	A read-only single value file which exists on all cgroups.
1618
1619	The number of processes currently in the cgroup and its
1620	descendants.
1621
1622Organisational operations are not blocked by cgroup policies, so it is
1623possible to have pids.current > pids.max.  This can be done by either
1624setting the limit to be smaller than pids.current, or attaching enough
1625processes to the cgroup such that pids.current is larger than
1626pids.max.  However, it is not possible to violate a cgroup PID policy
1627through fork() or clone(). These will return -EAGAIN if the creation
1628of a new process would cause a cgroup policy to be violated.
1629
1630
1631Cpuset
1632------
1633
1634The "cpuset" controller provides a mechanism for constraining
1635the CPU and memory node placement of tasks to only the resources
1636specified in the cpuset interface files in a task's current cgroup.
1637This is especially valuable on large NUMA systems where placing jobs
1638on properly sized subsets of the systems with careful processor and
1639memory placement to reduce cross-node memory access and contention
1640can improve overall system performance.
1641
1642The "cpuset" controller is hierarchical.  That means the controller
1643cannot use CPUs or memory nodes not allowed in its parent.
1644
1645
1646Cpuset Interface Files
1647~~~~~~~~~~~~~~~~~~~~~~
1648
1649  cpuset.cpus
1650	A read-write multiple values file which exists on non-root
1651	cpuset-enabled cgroups.
1652
1653	It lists the requested CPUs to be used by tasks within this
1654	cgroup.  The actual list of CPUs to be granted, however, is
1655	subjected to constraints imposed by its parent and can differ
1656	from the requested CPUs.
1657
1658	The CPU numbers are comma-separated numbers or ranges.
1659	For example:
1660
1661	  # cat cpuset.cpus
1662	  0-4,6,8-10
1663
1664	An empty value indicates that the cgroup is using the same
1665	setting as the nearest cgroup ancestor with a non-empty
1666	"cpuset.cpus" or all the available CPUs if none is found.
1667
1668	The value of "cpuset.cpus" stays constant until the next update
1669	and won't be affected by any CPU hotplug events.
1670
1671  cpuset.cpus.effective
1672	A read-only multiple values file which exists on all
1673	cpuset-enabled cgroups.
1674
1675	It lists the onlined CPUs that are actually granted to this
1676	cgroup by its parent.  These CPUs are allowed to be used by
1677	tasks within the current cgroup.
1678
1679	If "cpuset.cpus" is empty, the "cpuset.cpus.effective" file shows
1680	all the CPUs from the parent cgroup that can be available to
1681	be used by this cgroup.  Otherwise, it should be a subset of
1682	"cpuset.cpus" unless none of the CPUs listed in "cpuset.cpus"
1683	can be granted.  In this case, it will be treated just like an
1684	empty "cpuset.cpus".
1685
1686	Its value will be affected by CPU hotplug events.
1687
1688  cpuset.mems
1689	A read-write multiple values file which exists on non-root
1690	cpuset-enabled cgroups.
1691
1692	It lists the requested memory nodes to be used by tasks within
1693	this cgroup.  The actual list of memory nodes granted, however,
1694	is subjected to constraints imposed by its parent and can differ
1695	from the requested memory nodes.
1696
1697	The memory node numbers are comma-separated numbers or ranges.
1698	For example:
1699
1700	  # cat cpuset.mems
1701	  0-1,3
1702
1703	An empty value indicates that the cgroup is using the same
1704	setting as the nearest cgroup ancestor with a non-empty
1705	"cpuset.mems" or all the available memory nodes if none
1706	is found.
1707
1708	The value of "cpuset.mems" stays constant until the next update
1709	and won't be affected by any memory nodes hotplug events.
1710
1711  cpuset.mems.effective
1712	A read-only multiple values file which exists on all
1713	cpuset-enabled cgroups.
1714
1715	It lists the onlined memory nodes that are actually granted to
1716	this cgroup by its parent. These memory nodes are allowed to
1717	be used by tasks within the current cgroup.
1718
1719	If "cpuset.mems" is empty, it shows all the memory nodes from the
1720	parent cgroup that will be available to be used by this cgroup.
1721	Otherwise, it should be a subset of "cpuset.mems" unless none of
1722	the memory nodes listed in "cpuset.mems" can be granted.  In this
1723	case, it will be treated just like an empty "cpuset.mems".
1724
1725	Its value will be affected by memory nodes hotplug events.
1726
1727  cpuset.cpus.partition
1728	A read-write single value file which exists on non-root
1729	cpuset-enabled cgroups.  This flag is owned by the parent cgroup
1730	and is not delegatable.
1731
1732        It accepts only the following input values when written to.
1733
1734        "root"   - a paritition root
1735        "member" - a non-root member of a partition
1736
1737	When set to be a partition root, the current cgroup is the
1738	root of a new partition or scheduling domain that comprises
1739	itself and all its descendants except those that are separate
1740	partition roots themselves and their descendants.  The root
1741	cgroup is always a partition root.
1742
1743	There are constraints on where a partition root can be set.
1744	It can only be set in a cgroup if all the following conditions
1745	are true.
1746
1747	1) The "cpuset.cpus" is not empty and the list of CPUs are
1748	   exclusive, i.e. they are not shared by any of its siblings.
1749	2) The parent cgroup is a partition root.
1750	3) The "cpuset.cpus" is also a proper subset of the parent's
1751	   "cpuset.cpus.effective".
1752	4) There is no child cgroups with cpuset enabled.  This is for
1753	   eliminating corner cases that have to be handled if such a
1754	   condition is allowed.
1755
1756	Setting it to partition root will take the CPUs away from the
1757	effective CPUs of the parent cgroup.  Once it is set, this
1758	file cannot be reverted back to "member" if there are any child
1759	cgroups with cpuset enabled.
1760
1761	A parent partition cannot distribute all its CPUs to its
1762	child partitions.  There must be at least one cpu left in the
1763	parent partition.
1764
1765	Once becoming a partition root, changes to "cpuset.cpus" is
1766	generally allowed as long as the first condition above is true,
1767	the change will not take away all the CPUs from the parent
1768	partition and the new "cpuset.cpus" value is a superset of its
1769	children's "cpuset.cpus" values.
1770
1771	Sometimes, external factors like changes to ancestors'
1772	"cpuset.cpus" or cpu hotplug can cause the state of the partition
1773	root to change.  On read, the "cpuset.sched.partition" file
1774	can show the following values.
1775
1776	"member"       Non-root member of a partition
1777	"root"         Partition root
1778	"root invalid" Invalid partition root
1779
1780	It is a partition root if the first 2 partition root conditions
1781	above are true and at least one CPU from "cpuset.cpus" is
1782	granted by the parent cgroup.
1783
1784	A partition root can become invalid if none of CPUs requested
1785	in "cpuset.cpus" can be granted by the parent cgroup or the
1786	parent cgroup is no longer a partition root itself.  In this
1787	case, it is not a real partition even though the restriction
1788	of the first partition root condition above will still apply.
1789	The cpu affinity of all the tasks in the cgroup will then be
1790	associated with CPUs in the nearest ancestor partition.
1791
1792	An invalid partition root can be transitioned back to a
1793	real partition root if at least one of the requested CPUs
1794	can now be granted by its parent.  In this case, the cpu
1795	affinity of all the tasks in the formerly invalid partition
1796	will be associated to the CPUs of the newly formed partition.
1797	Changing the partition state of an invalid partition root to
1798	"member" is always allowed even if child cpusets are present.
1799
1800
1801Device controller
1802-----------------
1803
1804Device controller manages access to device files. It includes both
1805creation of new device files (using mknod), and access to the
1806existing device files.
1807
1808Cgroup v2 device controller has no interface files and is implemented
1809on top of cgroup BPF. To control access to device files, a user may
1810create bpf programs of the BPF_CGROUP_DEVICE type and attach them
1811to cgroups. On an attempt to access a device file, corresponding
1812BPF programs will be executed, and depending on the return value
1813the attempt will succeed or fail with -EPERM.
1814
1815A BPF_CGROUP_DEVICE program takes a pointer to the bpf_cgroup_dev_ctx
1816structure, which describes the device access attempt: access type
1817(mknod/read/write) and device (type, major and minor numbers).
1818If the program returns 0, the attempt fails with -EPERM, otherwise
1819it succeeds.
1820
1821An example of BPF_CGROUP_DEVICE program may be found in the kernel
1822source tree in the tools/testing/selftests/bpf/dev_cgroup.c file.
1823
1824
1825RDMA
1826----
1827
1828The "rdma" controller regulates the distribution and accounting of
1829of RDMA resources.
1830
1831RDMA Interface Files
1832~~~~~~~~~~~~~~~~~~~~
1833
1834  rdma.max
1835	A readwrite nested-keyed file that exists for all the cgroups
1836	except root that describes current configured resource limit
1837	for a RDMA/IB device.
1838
1839	Lines are keyed by device name and are not ordered.
1840	Each line contains space separated resource name and its configured
1841	limit that can be distributed.
1842
1843	The following nested keys are defined.
1844
1845	  ==========	=============================
1846	  hca_handle	Maximum number of HCA Handles
1847	  hca_object 	Maximum number of HCA Objects
1848	  ==========	=============================
1849
1850	An example for mlx4 and ocrdma device follows::
1851
1852	  mlx4_0 hca_handle=2 hca_object=2000
1853	  ocrdma1 hca_handle=3 hca_object=max
1854
1855  rdma.current
1856	A read-only file that describes current resource usage.
1857	It exists for all the cgroup except root.
1858
1859	An example for mlx4 and ocrdma device follows::
1860
1861	  mlx4_0 hca_handle=1 hca_object=20
1862	  ocrdma1 hca_handle=1 hca_object=23
1863
1864
1865Misc
1866----
1867
1868perf_event
1869~~~~~~~~~~
1870
1871perf_event controller, if not mounted on a legacy hierarchy, is
1872automatically enabled on the v2 hierarchy so that perf events can
1873always be filtered by cgroup v2 path.  The controller can still be
1874moved to a legacy hierarchy after v2 hierarchy is populated.
1875
1876
1877Non-normative information
1878-------------------------
1879
1880This section contains information that isn't considered to be a part of
1881the stable kernel API and so is subject to change.
1882
1883
1884CPU controller root cgroup process behaviour
1885~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1886
1887When distributing CPU cycles in the root cgroup each thread in this
1888cgroup is treated as if it was hosted in a separate child cgroup of the
1889root cgroup. This child cgroup weight is dependent on its thread nice
1890level.
1891
1892For details of this mapping see sched_prio_to_weight array in
1893kernel/sched/core.c file (values from this array should be scaled
1894appropriately so the neutral - nice 0 - value is 100 instead of 1024).
1895
1896
1897IO controller root cgroup process behaviour
1898~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1899
1900Root cgroup processes are hosted in an implicit leaf child node.
1901When distributing IO resources this implicit child node is taken into
1902account as if it was a normal child cgroup of the root cgroup with a
1903weight value of 200.
1904
1905
1906Namespace
1907=========
1908
1909Basics
1910------
1911
1912cgroup namespace provides a mechanism to virtualize the view of the
1913"/proc/$PID/cgroup" file and cgroup mounts.  The CLONE_NEWCGROUP clone
1914flag can be used with clone(2) and unshare(2) to create a new cgroup
1915namespace.  The process running inside the cgroup namespace will have
1916its "/proc/$PID/cgroup" output restricted to cgroupns root.  The
1917cgroupns root is the cgroup of the process at the time of creation of
1918the cgroup namespace.
1919
1920Without cgroup namespace, the "/proc/$PID/cgroup" file shows the
1921complete path of the cgroup of a process.  In a container setup where
1922a set of cgroups and namespaces are intended to isolate processes the
1923"/proc/$PID/cgroup" file may leak potential system level information
1924to the isolated processes.  For Example::
1925
1926  # cat /proc/self/cgroup
1927  0::/batchjobs/container_id1
1928
1929The path '/batchjobs/container_id1' can be considered as system-data
1930and undesirable to expose to the isolated processes.  cgroup namespace
1931can be used to restrict visibility of this path.  For example, before
1932creating a cgroup namespace, one would see::
1933
1934  # ls -l /proc/self/ns/cgroup
1935  lrwxrwxrwx 1 root root 0 2014-07-15 10:37 /proc/self/ns/cgroup -> cgroup:[4026531835]
1936  # cat /proc/self/cgroup
1937  0::/batchjobs/container_id1
1938
1939After unsharing a new namespace, the view changes::
1940
1941  # ls -l /proc/self/ns/cgroup
1942  lrwxrwxrwx 1 root root 0 2014-07-15 10:35 /proc/self/ns/cgroup -> cgroup:[4026532183]
1943  # cat /proc/self/cgroup
1944  0::/
1945
1946When some thread from a multi-threaded process unshares its cgroup
1947namespace, the new cgroupns gets applied to the entire process (all
1948the threads).  This is natural for the v2 hierarchy; however, for the
1949legacy hierarchies, this may be unexpected.
1950
1951A cgroup namespace is alive as long as there are processes inside or
1952mounts pinning it.  When the last usage goes away, the cgroup
1953namespace is destroyed.  The cgroupns root and the actual cgroups
1954remain.
1955
1956
1957The Root and Views
1958------------------
1959
1960The 'cgroupns root' for a cgroup namespace is the cgroup in which the
1961process calling unshare(2) is running.  For example, if a process in
1962/batchjobs/container_id1 cgroup calls unshare, cgroup
1963/batchjobs/container_id1 becomes the cgroupns root.  For the
1964init_cgroup_ns, this is the real root ('/') cgroup.
1965
1966The cgroupns root cgroup does not change even if the namespace creator
1967process later moves to a different cgroup::
1968
1969  # ~/unshare -c # unshare cgroupns in some cgroup
1970  # cat /proc/self/cgroup
1971  0::/
1972  # mkdir sub_cgrp_1
1973  # echo 0 > sub_cgrp_1/cgroup.procs
1974  # cat /proc/self/cgroup
1975  0::/sub_cgrp_1
1976
1977Each process gets its namespace-specific view of "/proc/$PID/cgroup"
1978
1979Processes running inside the cgroup namespace will be able to see
1980cgroup paths (in /proc/self/cgroup) only inside their root cgroup.
1981From within an unshared cgroupns::
1982
1983  # sleep 100000 &
1984  [1] 7353
1985  # echo 7353 > sub_cgrp_1/cgroup.procs
1986  # cat /proc/7353/cgroup
1987  0::/sub_cgrp_1
1988
1989From the initial cgroup namespace, the real cgroup path will be
1990visible::
1991
1992  $ cat /proc/7353/cgroup
1993  0::/batchjobs/container_id1/sub_cgrp_1
1994
1995From a sibling cgroup namespace (that is, a namespace rooted at a
1996different cgroup), the cgroup path relative to its own cgroup
1997namespace root will be shown.  For instance, if PID 7353's cgroup
1998namespace root is at '/batchjobs/container_id2', then it will see::
1999
2000  # cat /proc/7353/cgroup
2001  0::/../container_id2/sub_cgrp_1
2002
2003Note that the relative path always starts with '/' to indicate that
2004its relative to the cgroup namespace root of the caller.
2005
2006
2007Migration and setns(2)
2008----------------------
2009
2010Processes inside a cgroup namespace can move into and out of the
2011namespace root if they have proper access to external cgroups.  For
2012example, from inside a namespace with cgroupns root at
2013/batchjobs/container_id1, and assuming that the global hierarchy is
2014still accessible inside cgroupns::
2015
2016  # cat /proc/7353/cgroup
2017  0::/sub_cgrp_1
2018  # echo 7353 > batchjobs/container_id2/cgroup.procs
2019  # cat /proc/7353/cgroup
2020  0::/../container_id2
2021
2022Note that this kind of setup is not encouraged.  A task inside cgroup
2023namespace should only be exposed to its own cgroupns hierarchy.
2024
2025setns(2) to another cgroup namespace is allowed when:
2026
2027(a) the process has CAP_SYS_ADMIN against its current user namespace
2028(b) the process has CAP_SYS_ADMIN against the target cgroup
2029    namespace's userns
2030
2031No implicit cgroup changes happen with attaching to another cgroup
2032namespace.  It is expected that the someone moves the attaching
2033process under the target cgroup namespace root.
2034
2035
2036Interaction with Other Namespaces
2037---------------------------------
2038
2039Namespace specific cgroup hierarchy can be mounted by a process
2040running inside a non-init cgroup namespace::
2041
2042  # mount -t cgroup2 none $MOUNT_POINT
2043
2044This will mount the unified cgroup hierarchy with cgroupns root as the
2045filesystem root.  The process needs CAP_SYS_ADMIN against its user and
2046mount namespaces.
2047
2048The virtualization of /proc/self/cgroup file combined with restricting
2049the view of cgroup hierarchy by namespace-private cgroupfs mount
2050provides a properly isolated cgroup view inside the container.
2051
2052
2053Information on Kernel Programming
2054=================================
2055
2056This section contains kernel programming information in the areas
2057where interacting with cgroup is necessary.  cgroup core and
2058controllers are not covered.
2059
2060
2061Filesystem Support for Writeback
2062--------------------------------
2063
2064A filesystem can support cgroup writeback by updating
2065address_space_operations->writepage[s]() to annotate bio's using the
2066following two functions.
2067
2068  wbc_init_bio(@wbc, @bio)
2069	Should be called for each bio carrying writeback data and
2070	associates the bio with the inode's owner cgroup and the
2071	corresponding request queue.  This must be called after
2072	a queue (device) has been associated with the bio and
2073	before submission.
2074
2075  wbc_account_io(@wbc, @page, @bytes)
2076	Should be called for each data segment being written out.
2077	While this function doesn't care exactly when it's called
2078	during the writeback session, it's the easiest and most
2079	natural to call it as data segments are added to a bio.
2080
2081With writeback bio's annotated, cgroup support can be enabled per
2082super_block by setting SB_I_CGROUPWB in ->s_iflags.  This allows for
2083selective disabling of cgroup writeback support which is helpful when
2084certain filesystem features, e.g. journaled data mode, are
2085incompatible.
2086
2087wbc_init_bio() binds the specified bio to its cgroup.  Depending on
2088the configuration, the bio may be executed at a lower priority and if
2089the writeback session is holding shared resources, e.g. a journal
2090entry, may lead to priority inversion.  There is no one easy solution
2091for the problem.  Filesystems can try to work around specific problem
2092cases by skipping wbc_init_bio() and using bio_associate_blkg()
2093directly.
2094
2095
2096Deprecated v1 Core Features
2097===========================
2098
2099- Multiple hierarchies including named ones are not supported.
2100
2101- All v1 mount options are not supported.
2102
2103- The "tasks" file is removed and "cgroup.procs" is not sorted.
2104
2105- "cgroup.clone_children" is removed.
2106
2107- /proc/cgroups is meaningless for v2.  Use "cgroup.controllers" file
2108  at the root instead.
2109
2110
2111Issues with v1 and Rationales for v2
2112====================================
2113
2114Multiple Hierarchies
2115--------------------
2116
2117cgroup v1 allowed an arbitrary number of hierarchies and each
2118hierarchy could host any number of controllers.  While this seemed to
2119provide a high level of flexibility, it wasn't useful in practice.
2120
2121For example, as there is only one instance of each controller, utility
2122type controllers such as freezer which can be useful in all
2123hierarchies could only be used in one.  The issue is exacerbated by
2124the fact that controllers couldn't be moved to another hierarchy once
2125hierarchies were populated.  Another issue was that all controllers
2126bound to a hierarchy were forced to have exactly the same view of the
2127hierarchy.  It wasn't possible to vary the granularity depending on
2128the specific controller.
2129
2130In practice, these issues heavily limited which controllers could be
2131put on the same hierarchy and most configurations resorted to putting
2132each controller on its own hierarchy.  Only closely related ones, such
2133as the cpu and cpuacct controllers, made sense to be put on the same
2134hierarchy.  This often meant that userland ended up managing multiple
2135similar hierarchies repeating the same steps on each hierarchy
2136whenever a hierarchy management operation was necessary.
2137
2138Furthermore, support for multiple hierarchies came at a steep cost.
2139It greatly complicated cgroup core implementation but more importantly
2140the support for multiple hierarchies restricted how cgroup could be
2141used in general and what controllers was able to do.
2142
2143There was no limit on how many hierarchies there might be, which meant
2144that a thread's cgroup membership couldn't be described in finite
2145length.  The key might contain any number of entries and was unlimited
2146in length, which made it highly awkward to manipulate and led to
2147addition of controllers which existed only to identify membership,
2148which in turn exacerbated the original problem of proliferating number
2149of hierarchies.
2150
2151Also, as a controller couldn't have any expectation regarding the
2152topologies of hierarchies other controllers might be on, each
2153controller had to assume that all other controllers were attached to
2154completely orthogonal hierarchies.  This made it impossible, or at
2155least very cumbersome, for controllers to cooperate with each other.
2156
2157In most use cases, putting controllers on hierarchies which are
2158completely orthogonal to each other isn't necessary.  What usually is
2159called for is the ability to have differing levels of granularity
2160depending on the specific controller.  In other words, hierarchy may
2161be collapsed from leaf towards root when viewed from specific
2162controllers.  For example, a given configuration might not care about
2163how memory is distributed beyond a certain level while still wanting
2164to control how CPU cycles are distributed.
2165
2166
2167Thread Granularity
2168------------------
2169
2170cgroup v1 allowed threads of a process to belong to different cgroups.
2171This didn't make sense for some controllers and those controllers
2172ended up implementing different ways to ignore such situations but
2173much more importantly it blurred the line between API exposed to
2174individual applications and system management interface.
2175
2176Generally, in-process knowledge is available only to the process
2177itself; thus, unlike service-level organization of processes,
2178categorizing threads of a process requires active participation from
2179the application which owns the target process.
2180
2181cgroup v1 had an ambiguously defined delegation model which got abused
2182in combination with thread granularity.  cgroups were delegated to
2183individual applications so that they can create and manage their own
2184sub-hierarchies and control resource distributions along them.  This
2185effectively raised cgroup to the status of a syscall-like API exposed
2186to lay programs.
2187
2188First of all, cgroup has a fundamentally inadequate interface to be
2189exposed this way.  For a process to access its own knobs, it has to
2190extract the path on the target hierarchy from /proc/self/cgroup,
2191construct the path by appending the name of the knob to the path, open
2192and then read and/or write to it.  This is not only extremely clunky
2193and unusual but also inherently racy.  There is no conventional way to
2194define transaction across the required steps and nothing can guarantee
2195that the process would actually be operating on its own sub-hierarchy.
2196
2197cgroup controllers implemented a number of knobs which would never be
2198accepted as public APIs because they were just adding control knobs to
2199system-management pseudo filesystem.  cgroup ended up with interface
2200knobs which were not properly abstracted or refined and directly
2201revealed kernel internal details.  These knobs got exposed to
2202individual applications through the ill-defined delegation mechanism
2203effectively abusing cgroup as a shortcut to implementing public APIs
2204without going through the required scrutiny.
2205
2206This was painful for both userland and kernel.  Userland ended up with
2207misbehaving and poorly abstracted interfaces and kernel exposing and
2208locked into constructs inadvertently.
2209
2210
2211Competition Between Inner Nodes and Threads
2212-------------------------------------------
2213
2214cgroup v1 allowed threads to be in any cgroups which created an
2215interesting problem where threads belonging to a parent cgroup and its
2216children cgroups competed for resources.  This was nasty as two
2217different types of entities competed and there was no obvious way to
2218settle it.  Different controllers did different things.
2219
2220The cpu controller considered threads and cgroups as equivalents and
2221mapped nice levels to cgroup weights.  This worked for some cases but
2222fell flat when children wanted to be allocated specific ratios of CPU
2223cycles and the number of internal threads fluctuated - the ratios
2224constantly changed as the number of competing entities fluctuated.
2225There also were other issues.  The mapping from nice level to weight
2226wasn't obvious or universal, and there were various other knobs which
2227simply weren't available for threads.
2228
2229The io controller implicitly created a hidden leaf node for each
2230cgroup to host the threads.  The hidden leaf had its own copies of all
2231the knobs with ``leaf_`` prefixed.  While this allowed equivalent
2232control over internal threads, it was with serious drawbacks.  It
2233always added an extra layer of nesting which wouldn't be necessary
2234otherwise, made the interface messy and significantly complicated the
2235implementation.
2236
2237The memory controller didn't have a way to control what happened
2238between internal tasks and child cgroups and the behavior was not
2239clearly defined.  There were attempts to add ad-hoc behaviors and
2240knobs to tailor the behavior to specific workloads which would have
2241led to problems extremely difficult to resolve in the long term.
2242
2243Multiple controllers struggled with internal tasks and came up with
2244different ways to deal with it; unfortunately, all the approaches were
2245severely flawed and, furthermore, the widely different behaviors
2246made cgroup as a whole highly inconsistent.
2247
2248This clearly is a problem which needs to be addressed from cgroup core
2249in a uniform way.
2250
2251
2252Other Interface Issues
2253----------------------
2254
2255cgroup v1 grew without oversight and developed a large number of
2256idiosyncrasies and inconsistencies.  One issue on the cgroup core side
2257was how an empty cgroup was notified - a userland helper binary was
2258forked and executed for each event.  The event delivery wasn't
2259recursive or delegatable.  The limitations of the mechanism also led
2260to in-kernel event delivery filtering mechanism further complicating
2261the interface.
2262
2263Controller interfaces were problematic too.  An extreme example is
2264controllers completely ignoring hierarchical organization and treating
2265all cgroups as if they were all located directly under the root
2266cgroup.  Some controllers exposed a large amount of inconsistent
2267implementation details to userland.
2268
2269There also was no consistency across controllers.  When a new cgroup
2270was created, some controllers defaulted to not imposing extra
2271restrictions while others disallowed any resource usage until
2272explicitly configured.  Configuration knobs for the same type of
2273control used widely differing naming schemes and formats.  Statistics
2274and information knobs were named arbitrarily and used different
2275formats and units even in the same controller.
2276
2277cgroup v2 establishes common conventions where appropriate and updates
2278controllers so that they expose minimal and consistent interfaces.
2279
2280
2281Controller Issues and Remedies
2282------------------------------
2283
2284Memory
2285~~~~~~
2286
2287The original lower boundary, the soft limit, is defined as a limit
2288that is per default unset.  As a result, the set of cgroups that
2289global reclaim prefers is opt-in, rather than opt-out.  The costs for
2290optimizing these mostly negative lookups are so high that the
2291implementation, despite its enormous size, does not even provide the
2292basic desirable behavior.  First off, the soft limit has no
2293hierarchical meaning.  All configured groups are organized in a global
2294rbtree and treated like equal peers, regardless where they are located
2295in the hierarchy.  This makes subtree delegation impossible.  Second,
2296the soft limit reclaim pass is so aggressive that it not just
2297introduces high allocation latencies into the system, but also impacts
2298system performance due to overreclaim, to the point where the feature
2299becomes self-defeating.
2300
2301The memory.low boundary on the other hand is a top-down allocated
2302reserve.  A cgroup enjoys reclaim protection when it's within its low,
2303which makes delegation of subtrees possible.
2304
2305The original high boundary, the hard limit, is defined as a strict
2306limit that can not budge, even if the OOM killer has to be called.
2307But this generally goes against the goal of making the most out of the
2308available memory.  The memory consumption of workloads varies during
2309runtime, and that requires users to overcommit.  But doing that with a
2310strict upper limit requires either a fairly accurate prediction of the
2311working set size or adding slack to the limit.  Since working set size
2312estimation is hard and error prone, and getting it wrong results in
2313OOM kills, most users tend to err on the side of a looser limit and
2314end up wasting precious resources.
2315
2316The memory.high boundary on the other hand can be set much more
2317conservatively.  When hit, it throttles allocations by forcing them
2318into direct reclaim to work off the excess, but it never invokes the
2319OOM killer.  As a result, a high boundary that is chosen too
2320aggressively will not terminate the processes, but instead it will
2321lead to gradual performance degradation.  The user can monitor this
2322and make corrections until the minimal memory footprint that still
2323gives acceptable performance is found.
2324
2325In extreme cases, with many concurrent allocations and a complete
2326breakdown of reclaim progress within the group, the high boundary can
2327be exceeded.  But even then it's mostly better to satisfy the
2328allocation from the slack available in other groups or the rest of the
2329system than killing the group.  Otherwise, memory.max is there to
2330limit this type of spillover and ultimately contain buggy or even
2331malicious applications.
2332
2333Setting the original memory.limit_in_bytes below the current usage was
2334subject to a race condition, where concurrent charges could cause the
2335limit setting to fail. memory.max on the other hand will first set the
2336limit to prevent new charges, and then reclaim and OOM kill until the
2337new limit is met - or the task writing to memory.max is killed.
2338
2339The combined memory+swap accounting and limiting is replaced by real
2340control over swap space.
2341
2342The main argument for a combined memory+swap facility in the original
2343cgroup design was that global or parental pressure would always be
2344able to swap all anonymous memory of a child group, regardless of the
2345child's own (possibly untrusted) configuration.  However, untrusted
2346groups can sabotage swapping by other means - such as referencing its
2347anonymous memory in a tight loop - and an admin can not assume full
2348swappability when overcommitting untrusted jobs.
2349
2350For trusted jobs, on the other hand, a combined counter is not an
2351intuitive userspace interface, and it flies in the face of the idea
2352that cgroup controllers should account and limit specific physical
2353resources.  Swap space is a resource like all others in the system,
2354and that's why unified hierarchy allows distributing it separately.
2355