6 :Author: Tejun Heo <tj@kernel.org>
8 This is the authoritative documentation on the design, interface and
9 conventions of cgroup v2. It describes all userland-visible aspects
10 of cgroup including core and specific controller behaviors. All
11 future changes must be reflected in this document. Documentation for
12 v1 is available under Documentation/admin-guide/cgroup-v1/.
21 2-2. Organizing Processes and 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
30 2-5-1. Model of Delegation
31 2-5-2. Delegation Containment
33 2-6-1. Organize Once and Control
34 2-6-2. Avoid Name Collisions
35 3. Resource Distribution Models
43 4-3. Core Interface Files
46 5-1-1. CPU Interface Files
48 5-2-1. Memory Interface Files
49 5-2-2. Usage Guidelines
50 5-2-3. Memory Ownership
52 5-3-1. IO Interface Files
55 5-3-3-1. How IO Latency Throttling Works
56 5-3-3-2. IO Latency Interface Files
58 5-4-1. PID Interface Files
60 5.5-1. Cpuset Interface Files
63 5-7-1. RDMA Interface Files
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
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
92 "cgroup" stands for "control group" and is never capitalized. The
93 singular form is used to designate the whole feature and also as a
94 qualifier as in "cgroup controllers". When explicitly referring to
95 multiple individual control groups, the plural form "cgroups" is used.
101 cgroup is a mechanism to organize processes hierarchically and
102 distribute system resources along the hierarchy in a controlled and
105 cgroup is largely composed of two parts - the core and controllers.
106 cgroup core is primarily responsible for hierarchically organizing
107 processes. A cgroup controller is usually responsible for
108 distributing a specific type of system resource along the hierarchy
109 although there are utility controllers which serve purposes other than
110 resource distribution.
112 cgroups form a tree structure and every process in the system belongs
113 to one and only one cgroup. All threads of a process belong to the
114 same cgroup. On creation, all processes are put in the cgroup that
115 the parent process belongs to at the time. A process can be migrated
116 to another cgroup. Migration of a process doesn't affect already
117 existing descendant processes.
119 Following certain structural constraints, controllers may be enabled or
120 disabled selectively on a cgroup. All controller behaviors are
121 hierarchical - if a controller is enabled on a cgroup, it affects all
122 processes which belong to the cgroups consisting the inclusive
123 sub-hierarchy of the cgroup. When a controller is enabled on a nested
124 cgroup, it always restricts the resource distribution further. The
125 restrictions set closer to the root in the hierarchy can not be
126 overridden from further away.
135 Unlike v1, cgroup v2 has only single hierarchy. The cgroup v2
136 hierarchy can be mounted with the following mount command::
138 # mount -t cgroup2 none $MOUNT_POINT
140 cgroup2 filesystem has the magic number 0x63677270 ("cgrp"). All
141 controllers which support v2 and are not bound to a v1 hierarchy are
142 automatically bound to the v2 hierarchy and show up at the root.
143 Controllers which are not in active use in the v2 hierarchy can be
144 bound to other hierarchies. This allows mixing v2 hierarchy with the
145 legacy v1 multiple hierarchies in a fully backward compatible way.
147 A controller can be moved across hierarchies only after the controller
148 is no longer referenced in its current hierarchy. Because per-cgroup
149 controller states are destroyed asynchronously and controllers may
150 have lingering references, a controller may not show up immediately on
151 the v2 hierarchy after the final umount of the previous hierarchy.
152 Similarly, a controller should be fully disabled to be moved out of
153 the unified hierarchy and it may take some time for the disabled
154 controller to become available for other hierarchies; furthermore, due
155 to inter-controller dependencies, other controllers may need to be
158 While useful for development and manual configurations, moving
159 controllers dynamically between the v2 and other hierarchies is
160 strongly discouraged for production use. It is recommended to decide
161 the hierarchies and controller associations before starting using the
162 controllers after system boot.
164 During transition to v2, system management software might still
165 automount the v1 cgroup filesystem and so hijack all controllers
166 during boot, before manual intervention is possible. To make testing
167 and experimenting easier, the kernel parameter cgroup_no_v1= allows
168 disabling controllers in v1 and make them always available in v2.
170 cgroup v2 currently supports the following mount options.
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.
182 Only populate memory.events with data for the current cgroup,
183 and not any subtrees. This is legacy behaviour, the default
184 behaviour without this option is to include subtree counts.
185 This option is system wide and can only be set on mount or
186 modified through remount from the init namespace. The mount
187 option is ignored on non-init namespace mounts.
190 Organizing Processes and Threads
191 --------------------------------
196 Initially, only the root cgroup exists to which all processes belong.
197 A child cgroup can be created by creating a sub-directory::
201 A given cgroup may have multiple child cgroups forming a tree
202 structure. Each cgroup has a read-writable interface file
203 "cgroup.procs". When read, it lists the PIDs of all processes which
204 belong to the cgroup one-per-line. The PIDs are not ordered and the
205 same PID may show up more than once if the process got moved to
206 another cgroup and then back or the PID got recycled while reading.
208 A process can be migrated into a cgroup by writing its PID to the
209 target cgroup's "cgroup.procs" file. Only one process can be migrated
210 on a single write(2) call. If a process is composed of multiple
211 threads, writing the PID of any thread migrates all threads of the
214 When a process forks a child process, the new process is born into the
215 cgroup that the forking process belongs to at the time of the
216 operation. After exit, a process stays associated with the cgroup
217 that it belonged to at the time of exit until it's reaped; however, a
218 zombie process does not appear in "cgroup.procs" and thus can't be
219 moved to another cgroup.
221 A cgroup which doesn't have any children or live processes can be
222 destroyed by removing the directory. Note that a cgroup which doesn't
223 have any children and is associated only with zombie processes is
224 considered empty and can be removed::
228 "/proc/$PID/cgroup" lists a process's cgroup membership. If legacy
229 cgroup is in use in the system, this file may contain multiple lines,
230 one for each hierarchy. The entry for cgroup v2 is always in the
233 # cat /proc/842/cgroup
235 0::/test-cgroup/test-cgroup-nested
237 If the process becomes a zombie and the cgroup it was associated with
238 is removed subsequently, " (deleted)" is appended to the path::
240 # cat /proc/842/cgroup
242 0::/test-cgroup/test-cgroup-nested (deleted)
248 cgroup v2 supports thread granularity for a subset of controllers to
249 support use cases requiring hierarchical resource distribution across
250 the threads of a group of processes. By default, all threads of a
251 process belong to the same cgroup, which also serves as the resource
252 domain to host resource consumptions which are not specific to a
253 process or thread. The thread mode allows threads to be spread across
254 a subtree while still maintaining the common resource domain for them.
256 Controllers which support thread mode are called threaded controllers.
257 The ones which don't are called domain controllers.
259 Marking a cgroup threaded makes it join the resource domain of its
260 parent as a threaded cgroup. The parent may be another threaded
261 cgroup whose resource domain is further up in the hierarchy. The root
262 of a threaded subtree, that is, the nearest ancestor which is not
263 threaded, is called threaded domain or thread root interchangeably and
264 serves as the resource domain for the entire subtree.
266 Inside a threaded subtree, threads of a process can be put in
267 different cgroups and are not subject to the no internal process
268 constraint - threaded controllers can be enabled on non-leaf cgroups
269 whether they have threads in them or not.
271 As the threaded domain cgroup hosts all the domain resource
272 consumptions of the subtree, it is considered to have internal
273 resource consumptions whether there are processes in it or not and
274 can't have populated child cgroups which aren't threaded. Because the
275 root cgroup is not subject to no internal process constraint, it can
276 serve both as a threaded domain and a parent to domain cgroups.
278 The current operation mode or type of the cgroup is shown in the
279 "cgroup.type" file which indicates whether the cgroup is a normal
280 domain, a domain which is serving as the domain of a threaded subtree,
281 or a threaded cgroup.
283 On creation, a cgroup is always a domain cgroup and can be made
284 threaded by writing "threaded" to the "cgroup.type" file. The
285 operation is single direction::
287 # echo threaded > cgroup.type
289 Once threaded, the cgroup can't be made a domain again. To enable the
290 thread mode, the following conditions must be met.
292 - As the cgroup will join the parent's resource domain. The parent
293 must either be a valid (threaded) domain or a threaded cgroup.
295 - When the parent is an unthreaded domain, it must not have any domain
296 controllers enabled or populated domain children. The root is
297 exempt from this requirement.
299 Topology-wise, a cgroup can be in an invalid state. Please consider
300 the following topology::
302 A (threaded domain) - B (threaded) - C (domain, just created)
304 C is created as a domain but isn't connected to a parent which can
305 host child domains. C can't be used until it is turned into a
306 threaded cgroup. "cgroup.type" file will report "domain (invalid)" in
307 these cases. Operations which fail due to invalid topology use
308 EOPNOTSUPP as the errno.
310 A domain cgroup is turned into a threaded domain when one of its child
311 cgroup becomes threaded or threaded controllers are enabled in the
312 "cgroup.subtree_control" file while there are processes in the cgroup.
313 A threaded domain reverts to a normal domain when the conditions
316 When read, "cgroup.threads" contains the list of the thread IDs of all
317 threads in the cgroup. Except that the operations are per-thread
318 instead of per-process, "cgroup.threads" has the same format and
319 behaves the same way as "cgroup.procs". While "cgroup.threads" can be
320 written to in any cgroup, as it can only move threads inside the same
321 threaded domain, its operations are confined inside each threaded
324 The threaded domain cgroup serves as the resource domain for the whole
325 subtree, and, while the threads can be scattered across the subtree,
326 all the processes are considered to be in the threaded domain cgroup.
327 "cgroup.procs" in a threaded domain cgroup contains the PIDs of all
328 processes in the subtree and is not readable in the subtree proper.
329 However, "cgroup.procs" can be written to from anywhere in the subtree
330 to migrate all threads of the matching process to the cgroup.
332 Only threaded controllers can be enabled in a threaded subtree. When
333 a threaded controller is enabled inside a threaded subtree, it only
334 accounts for and controls resource consumptions associated with the
335 threads in the cgroup and its descendants. All consumptions which
336 aren't tied to a specific thread belong to the threaded domain cgroup.
338 Because a threaded subtree is exempt from no internal process
339 constraint, a threaded controller must be able to handle competition
340 between threads in a non-leaf cgroup and its child cgroups. Each
341 threaded controller defines how such competitions are handled.
344 [Un]populated Notification
345 --------------------------
347 Each non-root cgroup has a "cgroup.events" file which contains
348 "populated" field indicating whether the cgroup's sub-hierarchy has
349 live processes in it. Its value is 0 if there is no live process in
350 the cgroup and its descendants; otherwise, 1. poll and [id]notify
351 events are triggered when the value changes. This can be used, for
352 example, to start a clean-up operation after all processes of a given
353 sub-hierarchy have exited. The populated state updates and
354 notifications are recursive. Consider the following sub-hierarchy
355 where the numbers in the parentheses represent the numbers of processes
361 A, B and C's "populated" fields would be 1 while D's 0. After the one
362 process in C exits, B and C's "populated" fields would flip to "0" and
363 file modified events will be generated on the "cgroup.events" files of
367 Controlling Controllers
368 -----------------------
370 Enabling and Disabling
371 ~~~~~~~~~~~~~~~~~~~~~~
373 Each cgroup has a "cgroup.controllers" file which lists all
374 controllers available for the cgroup to enable::
376 # cat cgroup.controllers
379 No controller is enabled by default. Controllers can be enabled and
380 disabled by writing to the "cgroup.subtree_control" file::
382 # echo "+cpu +memory -io" > cgroup.subtree_control
384 Only controllers which are listed in "cgroup.controllers" can be
385 enabled. When multiple operations are specified as above, either they
386 all succeed or fail. If multiple operations on the same controller
387 are specified, the last one is effective.
389 Enabling a controller in a cgroup indicates that the distribution of
390 the target resource across its immediate children will be controlled.
391 Consider the following sub-hierarchy. The enabled controllers are
392 listed in parentheses::
394 A(cpu,memory) - B(memory) - C()
397 As A has "cpu" and "memory" enabled, A will control the distribution
398 of CPU cycles and memory to its children, in this case, B. As B has
399 "memory" enabled but not "CPU", C and D will compete freely on CPU
400 cycles but their division of memory available to B will be controlled.
402 As a controller regulates the distribution of the target resource to
403 the cgroup's children, enabling it creates the controller's interface
404 files in the child cgroups. In the above example, enabling "cpu" on B
405 would create the "cpu." prefixed controller interface files in C and
406 D. Likewise, disabling "memory" from B would remove the "memory."
407 prefixed controller interface files from C and D. This means that the
408 controller interface files - anything which doesn't start with
409 "cgroup." are owned by the parent rather than the cgroup itself.
415 Resources are distributed top-down and a cgroup can further distribute
416 a resource only if the resource has been distributed to it from the
417 parent. This means that all non-root "cgroup.subtree_control" files
418 can only contain controllers which are enabled in the parent's
419 "cgroup.subtree_control" file. A controller can be enabled only if
420 the parent has the controller enabled and a controller can't be
421 disabled if one or more children have it enabled.
424 No Internal Process Constraint
425 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
427 Non-root cgroups can distribute domain resources to their children
428 only when they don't have any processes of their own. In other words,
429 only domain cgroups which don't contain any processes can have domain
430 controllers enabled in their "cgroup.subtree_control" files.
432 This guarantees that, when a domain controller is looking at the part
433 of the hierarchy which has it enabled, processes are always only on
434 the leaves. This rules out situations where child cgroups compete
435 against internal processes of the parent.
437 The root cgroup is exempt from this restriction. Root contains
438 processes and anonymous resource consumption which can't be associated
439 with any other cgroups and requires special treatment from most
440 controllers. How resource consumption in the root cgroup is governed
441 is up to each controller (for more information on this topic please
442 refer to the Non-normative information section in the Controllers
445 Note that the restriction doesn't get in the way if there is no
446 enabled controller in the cgroup's "cgroup.subtree_control". This is
447 important as otherwise it wouldn't be possible to create children of a
448 populated cgroup. To control resource distribution of a cgroup, the
449 cgroup must create children and transfer all its processes to the
450 children before enabling controllers in its "cgroup.subtree_control"
460 A cgroup can be delegated in two ways. First, to a less privileged
461 user by granting write access of the directory and its "cgroup.procs",
462 "cgroup.threads" and "cgroup.subtree_control" files to the user.
463 Second, if the "nsdelegate" mount option is set, automatically to a
464 cgroup namespace on namespace creation.
466 Because the resource control interface files in a given directory
467 control the distribution of the parent's resources, the delegatee
468 shouldn't be allowed to write to them. For the first method, this is
469 achieved by not granting access to these files. For the second, the
470 kernel rejects writes to all files other than "cgroup.procs" and
471 "cgroup.subtree_control" on a namespace root from inside the
474 The end results are equivalent for both delegation types. Once
475 delegated, the user can build sub-hierarchy under the directory,
476 organize processes inside it as it sees fit and further distribute the
477 resources it received from the parent. The limits and other settings
478 of all resource controllers are hierarchical and regardless of what
479 happens in the delegated sub-hierarchy, nothing can escape the
480 resource restrictions imposed by the parent.
482 Currently, cgroup doesn't impose any restrictions on the number of
483 cgroups in or nesting depth of a delegated sub-hierarchy; however,
484 this may be limited explicitly in the future.
487 Delegation Containment
488 ~~~~~~~~~~~~~~~~~~~~~~
490 A delegated sub-hierarchy is contained in the sense that processes
491 can't be moved into or out of the sub-hierarchy by the delegatee.
493 For delegations to a less privileged user, this is achieved by
494 requiring the following conditions for a process with a non-root euid
495 to migrate a target process into a cgroup by writing its PID to the
498 - The writer must have write access to the "cgroup.procs" file.
500 - The writer must have write access to the "cgroup.procs" file of the
501 common ancestor of the source and destination cgroups.
503 The above two constraints ensure that while a delegatee may migrate
504 processes around freely in the delegated sub-hierarchy it can't pull
505 in from or push out to outside the sub-hierarchy.
507 For an example, let's assume cgroups C0 and C1 have been delegated to
508 user U0 who created C00, C01 under C0 and C10 under C1 as follows and
509 all processes under C0 and C1 belong to U0::
511 ~~~~~~~~~~~~~ - C0 - C00
514 ~~~~~~~~~~~~~ - C1 - C10
516 Let's also say U0 wants to write the PID of a process which is
517 currently in C10 into "C00/cgroup.procs". U0 has write access to the
518 file; however, the common ancestor of the source cgroup C10 and the
519 destination cgroup C00 is above the points of delegation and U0 would
520 not have write access to its "cgroup.procs" files and thus the write
521 will be denied with -EACCES.
523 For delegations to namespaces, containment is achieved by requiring
524 that both the source and destination cgroups are reachable from the
525 namespace of the process which is attempting the migration. If either
526 is not reachable, the migration is rejected with -ENOENT.
532 Organize Once and Control
533 ~~~~~~~~~~~~~~~~~~~~~~~~~
535 Migrating a process across cgroups is a relatively expensive operation
536 and stateful resources such as memory are not moved together with the
537 process. This is an explicit design decision as there often exist
538 inherent trade-offs between migration and various hot paths in terms
539 of synchronization cost.
541 As such, migrating processes across cgroups frequently as a means to
542 apply different resource restrictions is discouraged. A workload
543 should be assigned to a cgroup according to the system's logical and
544 resource structure once on start-up. Dynamic adjustments to resource
545 distribution can be made by changing controller configuration through
549 Avoid Name Collisions
550 ~~~~~~~~~~~~~~~~~~~~~
552 Interface files for a cgroup and its children cgroups occupy the same
553 directory and it is possible to create children cgroups which collide
554 with interface files.
556 All cgroup core interface files are prefixed with "cgroup." and each
557 controller's interface files are prefixed with the controller name and
558 a dot. A controller's name is composed of lower case alphabets and
559 '_'s but never begins with an '_' so it can be used as the prefix
560 character for collision avoidance. Also, interface file names won't
561 start or end with terms which are often used in categorizing workloads
562 such as job, service, slice, unit or workload.
564 cgroup doesn't do anything to prevent name collisions and it's the
565 user's responsibility to avoid them.
568 Resource Distribution Models
569 ============================
571 cgroup controllers implement several resource distribution schemes
572 depending on the resource type and expected use cases. This section
573 describes major schemes in use along with their expected behaviors.
579 A parent's resource is distributed by adding up the weights of all
580 active children and giving each the fraction matching the ratio of its
581 weight against the sum. As only children which can make use of the
582 resource at the moment participate in the distribution, this is
583 work-conserving. Due to the dynamic nature, this model is usually
584 used for stateless resources.
586 All weights are in the range [1, 10000] with the default at 100. This
587 allows symmetric multiplicative biases in both directions at fine
588 enough granularity while staying in the intuitive range.
590 As long as the weight is in range, all configuration combinations are
591 valid and there is no reason to reject configuration changes or
594 "cpu.weight" proportionally distributes CPU cycles to active children
595 and is an example of this type.
601 A child can only consume upto the configured amount of the resource.
602 Limits can be over-committed - the sum of the limits of children can
603 exceed the amount of resource available to the parent.
605 Limits are in the range [0, max] and defaults to "max", which is noop.
607 As limits can be over-committed, all configuration combinations are
608 valid and there is no reason to reject configuration changes or
611 "io.max" limits the maximum BPS and/or IOPS that a cgroup can consume
612 on an IO device and is an example of this type.
618 A cgroup is protected upto the configured amount of the resource
619 as long as the usages of all its ancestors are under their
620 protected levels. Protections can be hard guarantees or best effort
621 soft boundaries. Protections can also be over-committed in which case
622 only upto the amount available to the parent is protected among
625 Protections are in the range [0, max] and defaults to 0, which is
628 As protections can be over-committed, all configuration combinations
629 are valid and there is no reason to reject configuration changes or
632 "memory.low" implements best-effort memory protection and is an
633 example of this type.
639 A cgroup is exclusively allocated a certain amount of a finite
640 resource. Allocations can't be over-committed - the sum of the
641 allocations of children can not exceed the amount of resource
642 available to the parent.
644 Allocations are in the range [0, max] and defaults to 0, which is no
647 As allocations can't be over-committed, some configuration
648 combinations are invalid and should be rejected. Also, if the
649 resource is mandatory for execution of processes, process migrations
652 "cpu.rt.max" hard-allocates realtime slices and is an example of this
662 All interface files should be in one of the following formats whenever
665 New-line separated values
666 (when only one value can be written at once)
672 Space separated values
673 (when read-only or multiple values can be written at once)
685 KEY0 SUB_KEY0=VAL00 SUB_KEY1=VAL01...
686 KEY1 SUB_KEY0=VAL10 SUB_KEY1=VAL11...
689 For a writable file, the format for writing should generally match
690 reading; however, controllers may allow omitting later fields or
691 implement restricted shortcuts for most common use cases.
693 For both flat and nested keyed files, only the values for a single key
694 can be written at a time. For nested keyed files, the sub key pairs
695 may be specified in any order and not all pairs have to be specified.
701 - Settings for a single feature should be contained in a single file.
703 - The root cgroup should be exempt from resource control and thus
704 shouldn't have resource control interface files. Also,
705 informational files on the root cgroup which end up showing global
706 information available elsewhere shouldn't exist.
708 - The default time unit is microseconds. If a different unit is ever
709 used, an explicit unit suffix must be present.
711 - A parts-per quantity should use a percentage decimal with at least
712 two digit fractional part - e.g. 13.40.
714 - If a controller implements weight based resource distribution, its
715 interface file should be named "weight" and have the range [1,
716 10000] with 100 as the default. The values are chosen to allow
717 enough and symmetric bias in both directions while keeping it
718 intuitive (the default is 100%).
720 - If a controller implements an absolute resource guarantee and/or
721 limit, the interface files should be named "min" and "max"
722 respectively. If a controller implements best effort resource
723 guarantee and/or limit, the interface files should be named "low"
724 and "high" respectively.
726 In the above four control files, the special token "max" should be
727 used to represent upward infinity for both reading and writing.
729 - If a setting has a configurable default value and keyed specific
730 overrides, the default entry should be keyed with "default" and
731 appear as the first entry in the file.
733 The default value can be updated by writing either "default $VAL" or
736 When writing to update a specific override, "default" can be used as
737 the value to indicate removal of the override. Override entries
738 with "default" as the value must not appear when read.
740 For example, a setting which is keyed by major:minor device numbers
741 with integer values may look like the following::
743 # cat cgroup-example-interface-file
747 The default value can be updated by::
749 # echo 125 > cgroup-example-interface-file
753 # echo "default 125" > cgroup-example-interface-file
755 An override can be set by::
757 # echo "8:16 170" > cgroup-example-interface-file
761 # echo "8:0 default" > cgroup-example-interface-file
762 # cat cgroup-example-interface-file
766 - For events which are not very high frequency, an interface file
767 "events" should be created which lists event key value pairs.
768 Whenever a notifiable event happens, file modified event should be
769 generated on the file.
775 All cgroup core files are prefixed with "cgroup."
779 A read-write single value file which exists on non-root
782 When read, it indicates the current type of the cgroup, which
783 can be one of the following values.
785 - "domain" : A normal valid domain cgroup.
787 - "domain threaded" : A threaded domain cgroup which is
788 serving as the root of a threaded subtree.
790 - "domain invalid" : A cgroup which is in an invalid state.
791 It can't be populated or have controllers enabled. It may
792 be allowed to become a threaded cgroup.
794 - "threaded" : A threaded cgroup which is a member of a
797 A cgroup can be turned into a threaded cgroup by writing
798 "threaded" to this file.
801 A read-write new-line separated values file which exists on
804 When read, it lists the PIDs of all processes which belong to
805 the cgroup one-per-line. The PIDs are not ordered and the
806 same PID may show up more than once if the process got moved
807 to another cgroup and then back or the PID got recycled while
810 A PID can be written to migrate the process associated with
811 the PID to the cgroup. The writer should match all of the
812 following conditions.
814 - It must have write access to the "cgroup.procs" file.
816 - It must have write access to the "cgroup.procs" file of the
817 common ancestor of the source and destination cgroups.
819 When delegating a sub-hierarchy, write access to this file
820 should be granted along with the containing directory.
822 In a threaded cgroup, reading this file fails with EOPNOTSUPP
823 as all the processes belong to the thread root. Writing is
824 supported and moves every thread of the process to the cgroup.
827 A read-write new-line separated values file which exists on
830 When read, it lists the TIDs of all threads which belong to
831 the cgroup one-per-line. The TIDs are not ordered and the
832 same TID may show up more than once if the thread got moved to
833 another cgroup and then back or the TID got recycled while
836 A TID can be written to migrate the thread associated with the
837 TID to the cgroup. The writer should match all of the
838 following conditions.
840 - It must have write access to the "cgroup.threads" file.
842 - The cgroup that the thread is currently in must be in the
843 same resource domain as the destination cgroup.
845 - It must have write access to the "cgroup.procs" file of the
846 common ancestor of the source and destination cgroups.
848 When delegating a sub-hierarchy, write access to this file
849 should be granted along with the containing directory.
852 A read-only space separated values file which exists on all
855 It shows space separated list of all controllers available to
856 the cgroup. The controllers are not ordered.
858 cgroup.subtree_control
859 A read-write space separated values file which exists on all
860 cgroups. Starts out empty.
862 When read, it shows space separated list of the controllers
863 which are enabled to control resource distribution from the
864 cgroup to its children.
866 Space separated list of controllers prefixed with '+' or '-'
867 can be written to enable or disable controllers. A controller
868 name prefixed with '+' enables the controller and '-'
869 disables. If a controller appears more than once on the list,
870 the last one is effective. When multiple enable and disable
871 operations are specified, either all succeed or all fail.
874 A read-only flat-keyed file which exists on non-root cgroups.
875 The following entries are defined. Unless specified
876 otherwise, a value change in this file generates a file
880 1 if the cgroup or its descendants contains any live
881 processes; otherwise, 0.
883 1 if the cgroup is frozen; otherwise, 0.
885 cgroup.max.descendants
886 A read-write single value files. The default is "max".
888 Maximum allowed number of descent cgroups.
889 If the actual number of descendants is equal or larger,
890 an attempt to create a new cgroup in the hierarchy will fail.
893 A read-write single value files. The default is "max".
895 Maximum allowed descent depth below the current cgroup.
896 If the actual descent depth is equal or larger,
897 an attempt to create a new child cgroup will fail.
900 A read-only flat-keyed file with the following entries:
903 Total number of visible descendant cgroups.
906 Total number of dying descendant cgroups. A cgroup becomes
907 dying after being deleted by a user. The cgroup will remain
908 in dying state for some time undefined time (which can depend
909 on system load) before being completely destroyed.
911 A process can't enter a dying cgroup under any circumstances,
912 a dying cgroup can't revive.
914 A dying cgroup can consume system resources not exceeding
915 limits, which were active at the moment of cgroup deletion.
918 A read-write single value file which exists on non-root cgroups.
919 Allowed values are "0" and "1". The default is "0".
921 Writing "1" to the file causes freezing of the cgroup and all
922 descendant cgroups. This means that all belonging processes will
923 be stopped and will not run until the cgroup will be explicitly
924 unfrozen. Freezing of the cgroup may take some time; when this action
925 is completed, the "frozen" value in the cgroup.events control file
926 will be updated to "1" and the corresponding notification will be
929 A cgroup can be frozen either by its own settings, or by settings
930 of any ancestor cgroups. If any of ancestor cgroups is frozen, the
931 cgroup will remain frozen.
933 Processes in the frozen cgroup can be killed by a fatal signal.
934 They also can enter and leave a frozen cgroup: either by an explicit
935 move by a user, or if freezing of the cgroup races with fork().
936 If a process is moved to a frozen cgroup, it stops. If a process is
937 moved out of a frozen cgroup, it becomes running.
939 Frozen status of a cgroup doesn't affect any cgroup tree operations:
940 it's possible to delete a frozen (and empty) cgroup, as well as
941 create new sub-cgroups.
949 The "cpu" controllers regulates distribution of CPU cycles. This
950 controller implements weight and absolute bandwidth limit models for
951 normal scheduling policy and absolute bandwidth allocation model for
952 realtime scheduling policy.
954 In all the above models, cycles distribution is defined only on a temporal
955 base and it does not account for the frequency at which tasks are executed.
956 The (optional) utilization clamping support allows to hint the schedutil
957 cpufreq governor about the minimum desired frequency which should always be
958 provided by a CPU, as well as the maximum desired frequency, which should not
959 be exceeded by a CPU.
961 WARNING: cgroup2 doesn't yet support control of realtime processes and
962 the cpu controller can only be enabled when all RT processes are in
963 the root cgroup. Be aware that system management software may already
964 have placed RT processes into nonroot cgroups during the system boot
965 process, and these processes may need to be moved to the root cgroup
966 before the cpu controller can be enabled.
972 All time durations are in microseconds.
975 A read-only flat-keyed file which exists on non-root cgroups.
976 This file exists whether the controller is enabled or not.
978 It always reports the following three stats:
984 and the following three when the controller is enabled:
991 A read-write single value file which exists on non-root
992 cgroups. The default is "100".
994 The weight in the range [1, 10000].
997 A read-write single value file which exists on non-root
998 cgroups. The default is "0".
1000 The nice value is in the range [-20, 19].
1002 This interface file is an alternative interface for
1003 "cpu.weight" and allows reading and setting weight using the
1004 same values used by nice(2). Because the range is smaller and
1005 granularity is coarser for the nice values, the read value is
1006 the closest approximation of the current weight.
1009 A read-write two value file which exists on non-root cgroups.
1010 The default is "max 100000".
1012 The maximum bandwidth limit. It's in the following format::
1016 which indicates that the group may consume upto $MAX in each
1017 $PERIOD duration. "max" for $MAX indicates no limit. If only
1018 one number is written, $MAX is updated.
1021 A read-only nested-key file which exists on non-root cgroups.
1023 Shows pressure stall information for CPU. See
1024 Documentation/accounting/psi.rst for details.
1027 A read-write single value file which exists on non-root cgroups.
1028 The default is "0", i.e. no utilization boosting.
1030 The requested minimum utilization (protection) as a percentage
1031 rational number, e.g. 12.34 for 12.34%.
1033 This interface allows reading and setting minimum utilization clamp
1034 values similar to the sched_setattr(2). This minimum utilization
1035 value is used to clamp the task specific minimum utilization clamp.
1037 The requested minimum utilization (protection) is always capped by
1038 the current value for the maximum utilization (limit), i.e.
1042 A read-write single value file which exists on non-root cgroups.
1043 The default is "max". i.e. no utilization capping
1045 The requested maximum utilization (limit) as a percentage rational
1046 number, e.g. 98.76 for 98.76%.
1048 This interface allows reading and setting maximum utilization clamp
1049 values similar to the sched_setattr(2). This maximum utilization
1050 value is used to clamp the task specific maximum utilization clamp.
1057 The "memory" controller regulates distribution of memory. Memory is
1058 stateful and implements both limit and protection models. Due to the
1059 intertwining between memory usage and reclaim pressure and the
1060 stateful nature of memory, the distribution model is relatively
1063 While not completely water-tight, all major memory usages by a given
1064 cgroup are tracked so that the total memory consumption can be
1065 accounted and controlled to a reasonable extent. Currently, the
1066 following types of memory usages are tracked.
1068 - Userland memory - page cache and anonymous memory.
1070 - Kernel data structures such as dentries and inodes.
1072 - TCP socket buffers.
1074 The above list may expand in the future for better coverage.
1077 Memory Interface Files
1078 ~~~~~~~~~~~~~~~~~~~~~~
1080 All memory amounts are in bytes. If a value which is not aligned to
1081 PAGE_SIZE is written, the value may be rounded up to the closest
1082 PAGE_SIZE multiple when read back.
1085 A read-only single value file which exists on non-root
1088 The total amount of memory currently being used by the cgroup
1089 and its descendants.
1092 A read-write single value file which exists on non-root
1093 cgroups. The default is "0".
1095 Hard memory protection. If the memory usage of a cgroup
1096 is within its effective min boundary, the cgroup's memory
1097 won't be reclaimed under any conditions. If there is no
1098 unprotected reclaimable memory available, OOM killer
1099 is invoked. Above the effective min boundary (or
1100 effective low boundary if it is higher), pages are reclaimed
1101 proportionally to the overage, reducing reclaim pressure for
1104 Effective min boundary is limited by memory.min values of
1105 all ancestor cgroups. If there is memory.min overcommitment
1106 (child cgroup or cgroups are requiring more protected memory
1107 than parent will allow), then each child cgroup will get
1108 the part of parent's protection proportional to its
1109 actual memory usage below memory.min.
1111 Putting more memory than generally available under this
1112 protection is discouraged and may lead to constant OOMs.
1114 If a memory cgroup is not populated with processes,
1115 its memory.min is ignored.
1118 A read-write single value file which exists on non-root
1119 cgroups. The default is "0".
1121 Best-effort memory protection. If the memory usage of a
1122 cgroup is within its effective low boundary, the cgroup's
1123 memory won't be reclaimed unless there is no reclaimable
1124 memory available in unprotected cgroups.
1125 Above the effective low boundary (or
1126 effective min boundary if it is higher), pages are reclaimed
1127 proportionally to the overage, reducing reclaim pressure for
1130 Effective low boundary is limited by memory.low values of
1131 all ancestor cgroups. If there is memory.low overcommitment
1132 (child cgroup or cgroups are requiring more protected memory
1133 than parent will allow), then each child cgroup will get
1134 the part of parent's protection proportional to its
1135 actual memory usage below memory.low.
1137 Putting more memory than generally available under this
1138 protection is discouraged.
1141 A read-write single value file which exists on non-root
1142 cgroups. The default is "max".
1144 Memory usage throttle limit. This is the main mechanism to
1145 control memory usage of a cgroup. If a cgroup's usage goes
1146 over the high boundary, the processes of the cgroup are
1147 throttled and put under heavy reclaim pressure.
1149 Going over the high limit never invokes the OOM killer and
1150 under extreme conditions the limit may be breached.
1153 A read-write single value file which exists on non-root
1154 cgroups. The default is "max".
1156 Memory usage hard limit. This is the final protection
1157 mechanism. If a cgroup's memory usage reaches this limit and
1158 can't be reduced, the OOM killer is invoked in the cgroup.
1159 Under certain circumstances, the usage may go over the limit
1162 This is the ultimate protection mechanism. As long as the
1163 high limit is used and monitored properly, this limit's
1164 utility is limited to providing the final safety net.
1167 A read-write single value file which exists on non-root
1168 cgroups. The default value is "0".
1170 Determines whether the cgroup should be treated as
1171 an indivisible workload by the OOM killer. If set,
1172 all tasks belonging to the cgroup or to its descendants
1173 (if the memory cgroup is not a leaf cgroup) are killed
1174 together or not at all. This can be used to avoid
1175 partial kills to guarantee workload integrity.
1177 Tasks with the OOM protection (oom_score_adj set to -1000)
1178 are treated as an exception and are never killed.
1180 If the OOM killer is invoked in a cgroup, it's not going
1181 to kill any tasks outside of this cgroup, regardless
1182 memory.oom.group values of ancestor cgroups.
1185 A read-only flat-keyed file which exists on non-root cgroups.
1186 The following entries are defined. Unless specified
1187 otherwise, a value change in this file generates a file
1190 Note that all fields in this file are hierarchical and the
1191 file modified event can be generated due to an event down the
1192 hierarchy. For for the local events at the cgroup level see
1193 memory.events.local.
1196 The number of times the cgroup is reclaimed due to
1197 high memory pressure even though its usage is under
1198 the low boundary. This usually indicates that the low
1199 boundary is over-committed.
1202 The number of times processes of the cgroup are
1203 throttled and routed to perform direct memory reclaim
1204 because the high memory boundary was exceeded. For a
1205 cgroup whose memory usage is capped by the high limit
1206 rather than global memory pressure, this event's
1207 occurrences are expected.
1210 The number of times the cgroup's memory usage was
1211 about to go over the max boundary. If direct reclaim
1212 fails to bring it down, the cgroup goes to OOM state.
1215 The number of time the cgroup's memory usage was
1216 reached the limit and allocation was about to fail.
1218 Depending on context result could be invocation of OOM
1219 killer and retrying allocation or failing allocation.
1221 Failed allocation in its turn could be returned into
1222 userspace as -ENOMEM or silently ignored in cases like
1223 disk readahead. For now OOM in memory cgroup kills
1224 tasks iff shortage has happened inside page fault.
1226 This event is not raised if the OOM killer is not
1227 considered as an option, e.g. for failed high-order
1231 The number of processes belonging to this cgroup
1232 killed by any kind of OOM killer.
1235 Similar to memory.events but the fields in the file are local
1236 to the cgroup i.e. not hierarchical. The file modified event
1237 generated on this file reflects only the local events.
1240 A read-only flat-keyed file which exists on non-root cgroups.
1242 This breaks down the cgroup's memory footprint into different
1243 types of memory, type-specific details, and other information
1244 on the state and past events of the memory management system.
1246 All memory amounts are in bytes.
1248 The entries are ordered to be human readable, and new entries
1249 can show up in the middle. Don't rely on items remaining in a
1250 fixed position; use the keys to look up specific values!
1253 Amount of memory used in anonymous mappings such as
1254 brk(), sbrk(), and mmap(MAP_ANONYMOUS)
1257 Amount of memory used to cache filesystem data,
1258 including tmpfs and shared memory.
1261 Amount of memory allocated to kernel stacks.
1264 Amount of memory used for storing in-kernel data
1268 Amount of memory used in network transmission buffers
1271 Amount of cached filesystem data that is swap-backed,
1272 such as tmpfs, shm segments, shared anonymous mmap()s
1275 Amount of cached filesystem data mapped with mmap()
1278 Amount of cached filesystem data that was modified but
1279 not yet written back to disk
1282 Amount of cached filesystem data that was modified and
1283 is currently being written back to disk
1286 Amount of memory used in anonymous mappings backed by
1287 transparent hugepages
1289 inactive_anon, active_anon, inactive_file, active_file, unevictable
1290 Amount of memory, swap-backed and filesystem-backed,
1291 on the internal memory management lists used by the
1292 page reclaim algorithm.
1294 As these represent internal list state (eg. shmem pages are on anon
1295 memory management lists), inactive_foo + active_foo may not be equal to
1296 the value for the foo counter, since the foo counter is type-based, not
1300 Part of "slab" that might be reclaimed, such as
1301 dentries and inodes.
1304 Part of "slab" that cannot be reclaimed on memory
1308 Total number of page faults incurred
1311 Number of major page faults incurred
1315 Number of refaults of previously evicted pages
1319 Number of refaulted pages that were immediately activated
1321 workingset_nodereclaim
1323 Number of times a shadow node has been reclaimed
1327 Amount of scanned pages (in an active LRU list)
1331 Amount of scanned pages (in an inactive LRU list)
1335 Amount of reclaimed pages
1339 Amount of pages moved to the active LRU list
1343 Amount of pages moved to the inactive LRU list
1347 Amount of pages postponed to be freed under memory pressure
1351 Amount of reclaimed lazyfree pages
1355 Number of transparent hugepages which were allocated to satisfy
1356 a page fault, including COW faults. This counter is not present
1357 when CONFIG_TRANSPARENT_HUGEPAGE is not set.
1361 Number of transparent hugepages which were allocated to allow
1362 collapsing an existing range of pages. This counter is not
1363 present when CONFIG_TRANSPARENT_HUGEPAGE is not set.
1366 A read-only single value file which exists on non-root
1369 The total amount of swap currently being used by the cgroup
1370 and its descendants.
1373 A read-write single value file which exists on non-root
1374 cgroups. The default is "max".
1376 Swap usage hard limit. If a cgroup's swap usage reaches this
1377 limit, anonymous memory of the cgroup will not be swapped out.
1380 A read-only flat-keyed file which exists on non-root cgroups.
1381 The following entries are defined. Unless specified
1382 otherwise, a value change in this file generates a file
1386 The number of times the cgroup's swap usage was about
1387 to go over the max boundary and swap allocation
1391 The number of times swap allocation failed either
1392 because of running out of swap system-wide or max
1395 When reduced under the current usage, the existing swap
1396 entries are reclaimed gradually and the swap usage may stay
1397 higher than the limit for an extended period of time. This
1398 reduces the impact on the workload and memory management.
1401 A read-only nested-key file which exists on non-root cgroups.
1403 Shows pressure stall information for memory. See
1404 Documentation/accounting/psi.rst for details.
1410 "memory.high" is the main mechanism to control memory usage.
1411 Over-committing on high limit (sum of high limits > available memory)
1412 and letting global memory pressure to distribute memory according to
1413 usage is a viable strategy.
1415 Because breach of the high limit doesn't trigger the OOM killer but
1416 throttles the offending cgroup, a management agent has ample
1417 opportunities to monitor and take appropriate actions such as granting
1418 more memory or terminating the workload.
1420 Determining whether a cgroup has enough memory is not trivial as
1421 memory usage doesn't indicate whether the workload can benefit from
1422 more memory. For example, a workload which writes data received from
1423 network to a file can use all available memory but can also operate as
1424 performant with a small amount of memory. A measure of memory
1425 pressure - how much the workload is being impacted due to lack of
1426 memory - is necessary to determine whether a workload needs more
1427 memory; unfortunately, memory pressure monitoring mechanism isn't
1434 A memory area is charged to the cgroup which instantiated it and stays
1435 charged to the cgroup until the area is released. Migrating a process
1436 to a different cgroup doesn't move the memory usages that it
1437 instantiated while in the previous cgroup to the new cgroup.
1439 A memory area may be used by processes belonging to different cgroups.
1440 To which cgroup the area will be charged is in-deterministic; however,
1441 over time, the memory area is likely to end up in a cgroup which has
1442 enough memory allowance to avoid high reclaim pressure.
1444 If a cgroup sweeps a considerable amount of memory which is expected
1445 to be accessed repeatedly by other cgroups, it may make sense to use
1446 POSIX_FADV_DONTNEED to relinquish the ownership of memory areas
1447 belonging to the affected files to ensure correct memory ownership.
1453 The "io" controller regulates the distribution of IO resources. This
1454 controller implements both weight based and absolute bandwidth or IOPS
1455 limit distribution; however, weight based distribution is available
1456 only if cfq-iosched is in use and neither scheme is available for
1464 A read-only nested-keyed file which exists on non-root
1467 Lines are keyed by $MAJ:$MIN device numbers and not ordered.
1468 The following nested keys are defined.
1470 ====== =====================
1472 wbytes Bytes written
1473 rios Number of read IOs
1474 wios Number of write IOs
1475 dbytes Bytes discarded
1476 dios Number of discard IOs
1477 ====== =====================
1479 An example read output follows:
1481 8:16 rbytes=1459200 wbytes=314773504 rios=192 wios=353 dbytes=0 dios=0
1482 8:0 rbytes=90430464 wbytes=299008000 rios=8950 wios=1252 dbytes=50331648 dios=3021
1485 A read-write nested-keyed file with exists only on the root
1488 This file configures the Quality of Service of the IO cost
1489 model based controller (CONFIG_BLK_CGROUP_IOCOST) which
1490 currently implements "io.weight" proportional control. Lines
1491 are keyed by $MAJ:$MIN device numbers and not ordered. The
1492 line for a given device is populated on the first write for
1493 the device on "io.cost.qos" or "io.cost.model". The following
1494 nested keys are defined.
1496 ====== =====================================
1497 enable Weight-based control enable
1498 ctrl "auto" or "user"
1499 rpct Read latency percentile [0, 100]
1500 rlat Read latency threshold
1501 wpct Write latency percentile [0, 100]
1502 wlat Write latency threshold
1503 min Minimum scaling percentage [1, 10000]
1504 max Maximum scaling percentage [1, 10000]
1505 ====== =====================================
1507 The controller is disabled by default and can be enabled by
1508 setting "enable" to 1. "rpct" and "wpct" parameters default
1509 to zero and the controller uses internal device saturation
1510 state to adjust the overall IO rate between "min" and "max".
1512 When a better control quality is needed, latency QoS
1513 parameters can be configured. For example::
1515 8:16 enable=1 ctrl=auto rpct=95.00 rlat=75000 wpct=95.00 wlat=150000 min=50.00 max=150.0
1517 shows that on sdb, the controller is enabled, will consider
1518 the device saturated if the 95th percentile of read completion
1519 latencies is above 75ms or write 150ms, and adjust the overall
1520 IO issue rate between 50% and 150% accordingly.
1522 The lower the saturation point, the better the latency QoS at
1523 the cost of aggregate bandwidth. The narrower the allowed
1524 adjustment range between "min" and "max", the more conformant
1525 to the cost model the IO behavior. Note that the IO issue
1526 base rate may be far off from 100% and setting "min" and "max"
1527 blindly can lead to a significant loss of device capacity or
1528 control quality. "min" and "max" are useful for regulating
1529 devices which show wide temporary behavior changes - e.g. a
1530 ssd which accepts writes at the line speed for a while and
1531 then completely stalls for multiple seconds.
1533 When "ctrl" is "auto", the parameters are controlled by the
1534 kernel and may change automatically. Setting "ctrl" to "user"
1535 or setting any of the percentile and latency parameters puts
1536 it into "user" mode and disables the automatic changes. The
1537 automatic mode can be restored by setting "ctrl" to "auto".
1540 A read-write nested-keyed file with exists only on the root
1543 This file configures the cost model of the IO cost model based
1544 controller (CONFIG_BLK_CGROUP_IOCOST) which currently
1545 implements "io.weight" proportional control. Lines are keyed
1546 by $MAJ:$MIN device numbers and not ordered. The line for a
1547 given device is populated on the first write for the device on
1548 "io.cost.qos" or "io.cost.model". The following nested keys
1551 ===== ================================
1552 ctrl "auto" or "user"
1553 model The cost model in use - "linear"
1554 ===== ================================
1556 When "ctrl" is "auto", the kernel may change all parameters
1557 dynamically. When "ctrl" is set to "user" or any other
1558 parameters are written to, "ctrl" become "user" and the
1559 automatic changes are disabled.
1561 When "model" is "linear", the following model parameters are
1564 ============= ========================================
1565 [r|w]bps The maximum sequential IO throughput
1566 [r|w]seqiops The maximum 4k sequential IOs per second
1567 [r|w]randiops The maximum 4k random IOs per second
1568 ============= ========================================
1570 From the above, the builtin linear model determines the base
1571 costs of a sequential and random IO and the cost coefficient
1572 for the IO size. While simple, this model can cover most
1573 common device classes acceptably.
1575 The IO cost model isn't expected to be accurate in absolute
1576 sense and is scaled to the device behavior dynamically.
1578 If needed, tools/cgroup/iocost_coef_gen.py can be used to
1579 generate device-specific coefficients.
1582 A read-write flat-keyed file which exists on non-root cgroups.
1583 The default is "default 100".
1585 The first line is the default weight applied to devices
1586 without specific override. The rest are overrides keyed by
1587 $MAJ:$MIN device numbers and not ordered. The weights are in
1588 the range [1, 10000] and specifies the relative amount IO time
1589 the cgroup can use in relation to its siblings.
1591 The default weight can be updated by writing either "default
1592 $WEIGHT" or simply "$WEIGHT". Overrides can be set by writing
1593 "$MAJ:$MIN $WEIGHT" and unset by writing "$MAJ:$MIN default".
1595 An example read output follows::
1602 A read-write nested-keyed file which exists on non-root
1605 BPS and IOPS based IO limit. Lines are keyed by $MAJ:$MIN
1606 device numbers and not ordered. The following nested keys are
1609 ===== ==================================
1610 rbps Max read bytes per second
1611 wbps Max write bytes per second
1612 riops Max read IO operations per second
1613 wiops Max write IO operations per second
1614 ===== ==================================
1616 When writing, any number of nested key-value pairs can be
1617 specified in any order. "max" can be specified as the value
1618 to remove a specific limit. If the same key is specified
1619 multiple times, the outcome is undefined.
1621 BPS and IOPS are measured in each IO direction and IOs are
1622 delayed if limit is reached. Temporary bursts are allowed.
1624 Setting read limit at 2M BPS and write at 120 IOPS for 8:16::
1626 echo "8:16 rbps=2097152 wiops=120" > io.max
1628 Reading returns the following::
1630 8:16 rbps=2097152 wbps=max riops=max wiops=120
1632 Write IOPS limit can be removed by writing the following::
1634 echo "8:16 wiops=max" > io.max
1636 Reading now returns the following::
1638 8:16 rbps=2097152 wbps=max riops=max wiops=max
1641 A read-only nested-key file which exists on non-root cgroups.
1643 Shows pressure stall information for IO. See
1644 Documentation/accounting/psi.rst for details.
1650 Page cache is dirtied through buffered writes and shared mmaps and
1651 written asynchronously to the backing filesystem by the writeback
1652 mechanism. Writeback sits between the memory and IO domains and
1653 regulates the proportion of dirty memory by balancing dirtying and
1656 The io controller, in conjunction with the memory controller,
1657 implements control of page cache writeback IOs. The memory controller
1658 defines the memory domain that dirty memory ratio is calculated and
1659 maintained for and the io controller defines the io domain which
1660 writes out dirty pages for the memory domain. Both system-wide and
1661 per-cgroup dirty memory states are examined and the more restrictive
1662 of the two is enforced.
1664 cgroup writeback requires explicit support from the underlying
1665 filesystem. Currently, cgroup writeback is implemented on ext2, ext4
1666 and btrfs. On other filesystems, all writeback IOs are attributed to
1669 There are inherent differences in memory and writeback management
1670 which affects how cgroup ownership is tracked. Memory is tracked per
1671 page while writeback per inode. For the purpose of writeback, an
1672 inode is assigned to a cgroup and all IO requests to write dirty pages
1673 from the inode are attributed to that cgroup.
1675 As cgroup ownership for memory is tracked per page, there can be pages
1676 which are associated with different cgroups than the one the inode is
1677 associated with. These are called foreign pages. The writeback
1678 constantly keeps track of foreign pages and, if a particular foreign
1679 cgroup becomes the majority over a certain period of time, switches
1680 the ownership of the inode to that cgroup.
1682 While this model is enough for most use cases where a given inode is
1683 mostly dirtied by a single cgroup even when the main writing cgroup
1684 changes over time, use cases where multiple cgroups write to a single
1685 inode simultaneously are not supported well. In such circumstances, a
1686 significant portion of IOs are likely to be attributed incorrectly.
1687 As memory controller assigns page ownership on the first use and
1688 doesn't update it until the page is released, even if writeback
1689 strictly follows page ownership, multiple cgroups dirtying overlapping
1690 areas wouldn't work as expected. It's recommended to avoid such usage
1693 The sysctl knobs which affect writeback behavior are applied to cgroup
1694 writeback as follows.
1696 vm.dirty_background_ratio, vm.dirty_ratio
1697 These ratios apply the same to cgroup writeback with the
1698 amount of available memory capped by limits imposed by the
1699 memory controller and system-wide clean memory.
1701 vm.dirty_background_bytes, vm.dirty_bytes
1702 For cgroup writeback, this is calculated into ratio against
1703 total available memory and applied the same way as
1704 vm.dirty[_background]_ratio.
1710 This is a cgroup v2 controller for IO workload protection. You provide a group
1711 with a latency target, and if the average latency exceeds that target the
1712 controller will throttle any peers that have a lower latency target than the
1715 The limits are only applied at the peer level in the hierarchy. This means that
1716 in the diagram below, only groups A, B, and C will influence each other, and
1717 groups D and F will influence each other. Group G will influence nobody::
1726 So the ideal way to configure this is to set io.latency in groups A, B, and C.
1727 Generally you do not want to set a value lower than the latency your device
1728 supports. Experiment to find the value that works best for your workload.
1729 Start at higher than the expected latency for your device and watch the
1730 avg_lat value in io.stat for your workload group to get an idea of the
1731 latency you see during normal operation. Use the avg_lat value as a basis for
1732 your real setting, setting at 10-15% higher than the value in io.stat.
1734 How IO Latency Throttling Works
1735 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1737 io.latency is work conserving; so as long as everybody is meeting their latency
1738 target the controller doesn't do anything. Once a group starts missing its
1739 target it begins throttling any peer group that has a higher target than itself.
1740 This throttling takes 2 forms:
1742 - Queue depth throttling. This is the number of outstanding IO's a group is
1743 allowed to have. We will clamp down relatively quickly, starting at no limit
1744 and going all the way down to 1 IO at a time.
1746 - Artificial delay induction. There are certain types of IO that cannot be
1747 throttled without possibly adversely affecting higher priority groups. This
1748 includes swapping and metadata IO. These types of IO are allowed to occur
1749 normally, however they are "charged" to the originating group. If the
1750 originating group is being throttled you will see the use_delay and delay
1751 fields in io.stat increase. The delay value is how many microseconds that are
1752 being added to any process that runs in this group. Because this number can
1753 grow quite large if there is a lot of swapping or metadata IO occurring we
1754 limit the individual delay events to 1 second at a time.
1756 Once the victimized group starts meeting its latency target again it will start
1757 unthrottling any peer groups that were throttled previously. If the victimized
1758 group simply stops doing IO the global counter will unthrottle appropriately.
1760 IO Latency Interface Files
1761 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1764 This takes a similar format as the other controllers.
1766 "MAJOR:MINOR target=<target time in microseconds"
1769 If the controller is enabled you will see extra stats in io.stat in
1770 addition to the normal ones.
1773 This is the current queue depth for the group.
1776 This is an exponential moving average with a decay rate of 1/exp
1777 bound by the sampling interval. The decay rate interval can be
1778 calculated by multiplying the win value in io.stat by the
1779 corresponding number of samples based on the win value.
1782 The sampling window size in milliseconds. This is the minimum
1783 duration of time between evaluation events. Windows only elapse
1784 with IO activity. Idle periods extend the most recent window.
1789 The process number controller is used to allow a cgroup to stop any
1790 new tasks from being fork()'d or clone()'d after a specified limit is
1793 The number of tasks in a cgroup can be exhausted in ways which other
1794 controllers cannot prevent, thus warranting its own controller. For
1795 example, a fork bomb is likely to exhaust the number of tasks before
1796 hitting memory restrictions.
1798 Note that PIDs used in this controller refer to TIDs, process IDs as
1806 A read-write single value file which exists on non-root
1807 cgroups. The default is "max".
1809 Hard limit of number of processes.
1812 A read-only single value file which exists on all cgroups.
1814 The number of processes currently in the cgroup and its
1817 Organisational operations are not blocked by cgroup policies, so it is
1818 possible to have pids.current > pids.max. This can be done by either
1819 setting the limit to be smaller than pids.current, or attaching enough
1820 processes to the cgroup such that pids.current is larger than
1821 pids.max. However, it is not possible to violate a cgroup PID policy
1822 through fork() or clone(). These will return -EAGAIN if the creation
1823 of a new process would cause a cgroup policy to be violated.
1829 The "cpuset" controller provides a mechanism for constraining
1830 the CPU and memory node placement of tasks to only the resources
1831 specified in the cpuset interface files in a task's current cgroup.
1832 This is especially valuable on large NUMA systems where placing jobs
1833 on properly sized subsets of the systems with careful processor and
1834 memory placement to reduce cross-node memory access and contention
1835 can improve overall system performance.
1837 The "cpuset" controller is hierarchical. That means the controller
1838 cannot use CPUs or memory nodes not allowed in its parent.
1841 Cpuset Interface Files
1842 ~~~~~~~~~~~~~~~~~~~~~~
1845 A read-write multiple values file which exists on non-root
1846 cpuset-enabled cgroups.
1848 It lists the requested CPUs to be used by tasks within this
1849 cgroup. The actual list of CPUs to be granted, however, is
1850 subjected to constraints imposed by its parent and can differ
1851 from the requested CPUs.
1853 The CPU numbers are comma-separated numbers or ranges.
1859 An empty value indicates that the cgroup is using the same
1860 setting as the nearest cgroup ancestor with a non-empty
1861 "cpuset.cpus" or all the available CPUs if none is found.
1863 The value of "cpuset.cpus" stays constant until the next update
1864 and won't be affected by any CPU hotplug events.
1866 cpuset.cpus.effective
1867 A read-only multiple values file which exists on all
1868 cpuset-enabled cgroups.
1870 It lists the onlined CPUs that are actually granted to this
1871 cgroup by its parent. These CPUs are allowed to be used by
1872 tasks within the current cgroup.
1874 If "cpuset.cpus" is empty, the "cpuset.cpus.effective" file shows
1875 all the CPUs from the parent cgroup that can be available to
1876 be used by this cgroup. Otherwise, it should be a subset of
1877 "cpuset.cpus" unless none of the CPUs listed in "cpuset.cpus"
1878 can be granted. In this case, it will be treated just like an
1879 empty "cpuset.cpus".
1881 Its value will be affected by CPU hotplug events.
1884 A read-write multiple values file which exists on non-root
1885 cpuset-enabled cgroups.
1887 It lists the requested memory nodes to be used by tasks within
1888 this cgroup. The actual list of memory nodes granted, however,
1889 is subjected to constraints imposed by its parent and can differ
1890 from the requested memory nodes.
1892 The memory node numbers are comma-separated numbers or ranges.
1898 An empty value indicates that the cgroup is using the same
1899 setting as the nearest cgroup ancestor with a non-empty
1900 "cpuset.mems" or all the available memory nodes if none
1903 The value of "cpuset.mems" stays constant until the next update
1904 and won't be affected by any memory nodes hotplug events.
1906 cpuset.mems.effective
1907 A read-only multiple values file which exists on all
1908 cpuset-enabled cgroups.
1910 It lists the onlined memory nodes that are actually granted to
1911 this cgroup by its parent. These memory nodes are allowed to
1912 be used by tasks within the current cgroup.
1914 If "cpuset.mems" is empty, it shows all the memory nodes from the
1915 parent cgroup that will be available to be used by this cgroup.
1916 Otherwise, it should be a subset of "cpuset.mems" unless none of
1917 the memory nodes listed in "cpuset.mems" can be granted. In this
1918 case, it will be treated just like an empty "cpuset.mems".
1920 Its value will be affected by memory nodes hotplug events.
1922 cpuset.cpus.partition
1923 A read-write single value file which exists on non-root
1924 cpuset-enabled cgroups. This flag is owned by the parent cgroup
1925 and is not delegatable.
1927 It accepts only the following input values when written to.
1929 "root" - a partition root
1930 "member" - a non-root member of a partition
1932 When set to be a partition root, the current cgroup is the
1933 root of a new partition or scheduling domain that comprises
1934 itself and all its descendants except those that are separate
1935 partition roots themselves and their descendants. The root
1936 cgroup is always a partition root.
1938 There are constraints on where a partition root can be set.
1939 It can only be set in a cgroup if all the following conditions
1942 1) The "cpuset.cpus" is not empty and the list of CPUs are
1943 exclusive, i.e. they are not shared by any of its siblings.
1944 2) The parent cgroup is a partition root.
1945 3) The "cpuset.cpus" is also a proper subset of the parent's
1946 "cpuset.cpus.effective".
1947 4) There is no child cgroups with cpuset enabled. This is for
1948 eliminating corner cases that have to be handled if such a
1949 condition is allowed.
1951 Setting it to partition root will take the CPUs away from the
1952 effective CPUs of the parent cgroup. Once it is set, this
1953 file cannot be reverted back to "member" if there are any child
1954 cgroups with cpuset enabled.
1956 A parent partition cannot distribute all its CPUs to its
1957 child partitions. There must be at least one cpu left in the
1960 Once becoming a partition root, changes to "cpuset.cpus" is
1961 generally allowed as long as the first condition above is true,
1962 the change will not take away all the CPUs from the parent
1963 partition and the new "cpuset.cpus" value is a superset of its
1964 children's "cpuset.cpus" values.
1966 Sometimes, external factors like changes to ancestors'
1967 "cpuset.cpus" or cpu hotplug can cause the state of the partition
1968 root to change. On read, the "cpuset.sched.partition" file
1969 can show the following values.
1971 "member" Non-root member of a partition
1972 "root" Partition root
1973 "root invalid" Invalid partition root
1975 It is a partition root if the first 2 partition root conditions
1976 above are true and at least one CPU from "cpuset.cpus" is
1977 granted by the parent cgroup.
1979 A partition root can become invalid if none of CPUs requested
1980 in "cpuset.cpus" can be granted by the parent cgroup or the
1981 parent cgroup is no longer a partition root itself. In this
1982 case, it is not a real partition even though the restriction
1983 of the first partition root condition above will still apply.
1984 The cpu affinity of all the tasks in the cgroup will then be
1985 associated with CPUs in the nearest ancestor partition.
1987 An invalid partition root can be transitioned back to a
1988 real partition root if at least one of the requested CPUs
1989 can now be granted by its parent. In this case, the cpu
1990 affinity of all the tasks in the formerly invalid partition
1991 will be associated to the CPUs of the newly formed partition.
1992 Changing the partition state of an invalid partition root to
1993 "member" is always allowed even if child cpusets are present.
1999 Device controller manages access to device files. It includes both
2000 creation of new device files (using mknod), and access to the
2001 existing device files.
2003 Cgroup v2 device controller has no interface files and is implemented
2004 on top of cgroup BPF. To control access to device files, a user may
2005 create bpf programs of the BPF_CGROUP_DEVICE type and attach them
2006 to cgroups. On an attempt to access a device file, corresponding
2007 BPF programs will be executed, and depending on the return value
2008 the attempt will succeed or fail with -EPERM.
2010 A BPF_CGROUP_DEVICE program takes a pointer to the bpf_cgroup_dev_ctx
2011 structure, which describes the device access attempt: access type
2012 (mknod/read/write) and device (type, major and minor numbers).
2013 If the program returns 0, the attempt fails with -EPERM, otherwise
2016 An example of BPF_CGROUP_DEVICE program may be found in the kernel
2017 source tree in the tools/testing/selftests/bpf/dev_cgroup.c file.
2023 The "rdma" controller regulates the distribution and accounting of
2026 RDMA Interface Files
2027 ~~~~~~~~~~~~~~~~~~~~
2030 A readwrite nested-keyed file that exists for all the cgroups
2031 except root that describes current configured resource limit
2032 for a RDMA/IB device.
2034 Lines are keyed by device name and are not ordered.
2035 Each line contains space separated resource name and its configured
2036 limit that can be distributed.
2038 The following nested keys are defined.
2040 ========== =============================
2041 hca_handle Maximum number of HCA Handles
2042 hca_object Maximum number of HCA Objects
2043 ========== =============================
2045 An example for mlx4 and ocrdma device follows::
2047 mlx4_0 hca_handle=2 hca_object=2000
2048 ocrdma1 hca_handle=3 hca_object=max
2051 A read-only file that describes current resource usage.
2052 It exists for all the cgroup except root.
2054 An example for mlx4 and ocrdma device follows::
2056 mlx4_0 hca_handle=1 hca_object=20
2057 ocrdma1 hca_handle=1 hca_object=23
2066 perf_event controller, if not mounted on a legacy hierarchy, is
2067 automatically enabled on the v2 hierarchy so that perf events can
2068 always be filtered by cgroup v2 path. The controller can still be
2069 moved to a legacy hierarchy after v2 hierarchy is populated.
2072 Non-normative information
2073 -------------------------
2075 This section contains information that isn't considered to be a part of
2076 the stable kernel API and so is subject to change.
2079 CPU controller root cgroup process behaviour
2080 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2082 When distributing CPU cycles in the root cgroup each thread in this
2083 cgroup is treated as if it was hosted in a separate child cgroup of the
2084 root cgroup. This child cgroup weight is dependent on its thread nice
2087 For details of this mapping see sched_prio_to_weight array in
2088 kernel/sched/core.c file (values from this array should be scaled
2089 appropriately so the neutral - nice 0 - value is 100 instead of 1024).
2092 IO controller root cgroup process behaviour
2093 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2095 Root cgroup processes are hosted in an implicit leaf child node.
2096 When distributing IO resources this implicit child node is taken into
2097 account as if it was a normal child cgroup of the root cgroup with a
2098 weight value of 200.
2107 cgroup namespace provides a mechanism to virtualize the view of the
2108 "/proc/$PID/cgroup" file and cgroup mounts. The CLONE_NEWCGROUP clone
2109 flag can be used with clone(2) and unshare(2) to create a new cgroup
2110 namespace. The process running inside the cgroup namespace will have
2111 its "/proc/$PID/cgroup" output restricted to cgroupns root. The
2112 cgroupns root is the cgroup of the process at the time of creation of
2113 the cgroup namespace.
2115 Without cgroup namespace, the "/proc/$PID/cgroup" file shows the
2116 complete path of the cgroup of a process. In a container setup where
2117 a set of cgroups and namespaces are intended to isolate processes the
2118 "/proc/$PID/cgroup" file may leak potential system level information
2119 to the isolated processes. For Example::
2121 # cat /proc/self/cgroup
2122 0::/batchjobs/container_id1
2124 The path '/batchjobs/container_id1' can be considered as system-data
2125 and undesirable to expose to the isolated processes. cgroup namespace
2126 can be used to restrict visibility of this path. For example, before
2127 creating a cgroup namespace, one would see::
2129 # ls -l /proc/self/ns/cgroup
2130 lrwxrwxrwx 1 root root 0 2014-07-15 10:37 /proc/self/ns/cgroup -> cgroup:[4026531835]
2131 # cat /proc/self/cgroup
2132 0::/batchjobs/container_id1
2134 After unsharing a new namespace, the view changes::
2136 # ls -l /proc/self/ns/cgroup
2137 lrwxrwxrwx 1 root root 0 2014-07-15 10:35 /proc/self/ns/cgroup -> cgroup:[4026532183]
2138 # cat /proc/self/cgroup
2141 When some thread from a multi-threaded process unshares its cgroup
2142 namespace, the new cgroupns gets applied to the entire process (all
2143 the threads). This is natural for the v2 hierarchy; however, for the
2144 legacy hierarchies, this may be unexpected.
2146 A cgroup namespace is alive as long as there are processes inside or
2147 mounts pinning it. When the last usage goes away, the cgroup
2148 namespace is destroyed. The cgroupns root and the actual cgroups
2155 The 'cgroupns root' for a cgroup namespace is the cgroup in which the
2156 process calling unshare(2) is running. For example, if a process in
2157 /batchjobs/container_id1 cgroup calls unshare, cgroup
2158 /batchjobs/container_id1 becomes the cgroupns root. For the
2159 init_cgroup_ns, this is the real root ('/') cgroup.
2161 The cgroupns root cgroup does not change even if the namespace creator
2162 process later moves to a different cgroup::
2164 # ~/unshare -c # unshare cgroupns in some cgroup
2165 # cat /proc/self/cgroup
2168 # echo 0 > sub_cgrp_1/cgroup.procs
2169 # cat /proc/self/cgroup
2172 Each process gets its namespace-specific view of "/proc/$PID/cgroup"
2174 Processes running inside the cgroup namespace will be able to see
2175 cgroup paths (in /proc/self/cgroup) only inside their root cgroup.
2176 From within an unshared cgroupns::
2180 # echo 7353 > sub_cgrp_1/cgroup.procs
2181 # cat /proc/7353/cgroup
2184 From the initial cgroup namespace, the real cgroup path will be
2187 $ cat /proc/7353/cgroup
2188 0::/batchjobs/container_id1/sub_cgrp_1
2190 From a sibling cgroup namespace (that is, a namespace rooted at a
2191 different cgroup), the cgroup path relative to its own cgroup
2192 namespace root will be shown. For instance, if PID 7353's cgroup
2193 namespace root is at '/batchjobs/container_id2', then it will see::
2195 # cat /proc/7353/cgroup
2196 0::/../container_id2/sub_cgrp_1
2198 Note that the relative path always starts with '/' to indicate that
2199 its relative to the cgroup namespace root of the caller.
2202 Migration and setns(2)
2203 ----------------------
2205 Processes inside a cgroup namespace can move into and out of the
2206 namespace root if they have proper access to external cgroups. For
2207 example, from inside a namespace with cgroupns root at
2208 /batchjobs/container_id1, and assuming that the global hierarchy is
2209 still accessible inside cgroupns::
2211 # cat /proc/7353/cgroup
2213 # echo 7353 > batchjobs/container_id2/cgroup.procs
2214 # cat /proc/7353/cgroup
2215 0::/../container_id2
2217 Note that this kind of setup is not encouraged. A task inside cgroup
2218 namespace should only be exposed to its own cgroupns hierarchy.
2220 setns(2) to another cgroup namespace is allowed when:
2222 (a) the process has CAP_SYS_ADMIN against its current user namespace
2223 (b) the process has CAP_SYS_ADMIN against the target cgroup
2226 No implicit cgroup changes happen with attaching to another cgroup
2227 namespace. It is expected that the someone moves the attaching
2228 process under the target cgroup namespace root.
2231 Interaction with Other Namespaces
2232 ---------------------------------
2234 Namespace specific cgroup hierarchy can be mounted by a process
2235 running inside a non-init cgroup namespace::
2237 # mount -t cgroup2 none $MOUNT_POINT
2239 This will mount the unified cgroup hierarchy with cgroupns root as the
2240 filesystem root. The process needs CAP_SYS_ADMIN against its user and
2243 The virtualization of /proc/self/cgroup file combined with restricting
2244 the view of cgroup hierarchy by namespace-private cgroupfs mount
2245 provides a properly isolated cgroup view inside the container.
2248 Information on Kernel Programming
2249 =================================
2251 This section contains kernel programming information in the areas
2252 where interacting with cgroup is necessary. cgroup core and
2253 controllers are not covered.
2256 Filesystem Support for Writeback
2257 --------------------------------
2259 A filesystem can support cgroup writeback by updating
2260 address_space_operations->writepage[s]() to annotate bio's using the
2261 following two functions.
2263 wbc_init_bio(@wbc, @bio)
2264 Should be called for each bio carrying writeback data and
2265 associates the bio with the inode's owner cgroup and the
2266 corresponding request queue. This must be called after
2267 a queue (device) has been associated with the bio and
2270 wbc_account_cgroup_owner(@wbc, @page, @bytes)
2271 Should be called for each data segment being written out.
2272 While this function doesn't care exactly when it's called
2273 during the writeback session, it's the easiest and most
2274 natural to call it as data segments are added to a bio.
2276 With writeback bio's annotated, cgroup support can be enabled per
2277 super_block by setting SB_I_CGROUPWB in ->s_iflags. This allows for
2278 selective disabling of cgroup writeback support which is helpful when
2279 certain filesystem features, e.g. journaled data mode, are
2282 wbc_init_bio() binds the specified bio to its cgroup. Depending on
2283 the configuration, the bio may be executed at a lower priority and if
2284 the writeback session is holding shared resources, e.g. a journal
2285 entry, may lead to priority inversion. There is no one easy solution
2286 for the problem. Filesystems can try to work around specific problem
2287 cases by skipping wbc_init_bio() and using bio_associate_blkg()
2291 Deprecated v1 Core Features
2292 ===========================
2294 - Multiple hierarchies including named ones are not supported.
2296 - All v1 mount options are not supported.
2298 - The "tasks" file is removed and "cgroup.procs" is not sorted.
2300 - "cgroup.clone_children" is removed.
2302 - /proc/cgroups is meaningless for v2. Use "cgroup.controllers" file
2303 at the root instead.
2306 Issues with v1 and Rationales for v2
2307 ====================================
2309 Multiple Hierarchies
2310 --------------------
2312 cgroup v1 allowed an arbitrary number of hierarchies and each
2313 hierarchy could host any number of controllers. While this seemed to
2314 provide a high level of flexibility, it wasn't useful in practice.
2316 For example, as there is only one instance of each controller, utility
2317 type controllers such as freezer which can be useful in all
2318 hierarchies could only be used in one. The issue is exacerbated by
2319 the fact that controllers couldn't be moved to another hierarchy once
2320 hierarchies were populated. Another issue was that all controllers
2321 bound to a hierarchy were forced to have exactly the same view of the
2322 hierarchy. It wasn't possible to vary the granularity depending on
2323 the specific controller.
2325 In practice, these issues heavily limited which controllers could be
2326 put on the same hierarchy and most configurations resorted to putting
2327 each controller on its own hierarchy. Only closely related ones, such
2328 as the cpu and cpuacct controllers, made sense to be put on the same
2329 hierarchy. This often meant that userland ended up managing multiple
2330 similar hierarchies repeating the same steps on each hierarchy
2331 whenever a hierarchy management operation was necessary.
2333 Furthermore, support for multiple hierarchies came at a steep cost.
2334 It greatly complicated cgroup core implementation but more importantly
2335 the support for multiple hierarchies restricted how cgroup could be
2336 used in general and what controllers was able to do.
2338 There was no limit on how many hierarchies there might be, which meant
2339 that a thread's cgroup membership couldn't be described in finite
2340 length. The key might contain any number of entries and was unlimited
2341 in length, which made it highly awkward to manipulate and led to
2342 addition of controllers which existed only to identify membership,
2343 which in turn exacerbated the original problem of proliferating number
2346 Also, as a controller couldn't have any expectation regarding the
2347 topologies of hierarchies other controllers might be on, each
2348 controller had to assume that all other controllers were attached to
2349 completely orthogonal hierarchies. This made it impossible, or at
2350 least very cumbersome, for controllers to cooperate with each other.
2352 In most use cases, putting controllers on hierarchies which are
2353 completely orthogonal to each other isn't necessary. What usually is
2354 called for is the ability to have differing levels of granularity
2355 depending on the specific controller. In other words, hierarchy may
2356 be collapsed from leaf towards root when viewed from specific
2357 controllers. For example, a given configuration might not care about
2358 how memory is distributed beyond a certain level while still wanting
2359 to control how CPU cycles are distributed.
2365 cgroup v1 allowed threads of a process to belong to different cgroups.
2366 This didn't make sense for some controllers and those controllers
2367 ended up implementing different ways to ignore such situations but
2368 much more importantly it blurred the line between API exposed to
2369 individual applications and system management interface.
2371 Generally, in-process knowledge is available only to the process
2372 itself; thus, unlike service-level organization of processes,
2373 categorizing threads of a process requires active participation from
2374 the application which owns the target process.
2376 cgroup v1 had an ambiguously defined delegation model which got abused
2377 in combination with thread granularity. cgroups were delegated to
2378 individual applications so that they can create and manage their own
2379 sub-hierarchies and control resource distributions along them. This
2380 effectively raised cgroup to the status of a syscall-like API exposed
2383 First of all, cgroup has a fundamentally inadequate interface to be
2384 exposed this way. For a process to access its own knobs, it has to
2385 extract the path on the target hierarchy from /proc/self/cgroup,
2386 construct the path by appending the name of the knob to the path, open
2387 and then read and/or write to it. This is not only extremely clunky
2388 and unusual but also inherently racy. There is no conventional way to
2389 define transaction across the required steps and nothing can guarantee
2390 that the process would actually be operating on its own sub-hierarchy.
2392 cgroup controllers implemented a number of knobs which would never be
2393 accepted as public APIs because they were just adding control knobs to
2394 system-management pseudo filesystem. cgroup ended up with interface
2395 knobs which were not properly abstracted or refined and directly
2396 revealed kernel internal details. These knobs got exposed to
2397 individual applications through the ill-defined delegation mechanism
2398 effectively abusing cgroup as a shortcut to implementing public APIs
2399 without going through the required scrutiny.
2401 This was painful for both userland and kernel. Userland ended up with
2402 misbehaving and poorly abstracted interfaces and kernel exposing and
2403 locked into constructs inadvertently.
2406 Competition Between Inner Nodes and Threads
2407 -------------------------------------------
2409 cgroup v1 allowed threads to be in any cgroups which created an
2410 interesting problem where threads belonging to a parent cgroup and its
2411 children cgroups competed for resources. This was nasty as two
2412 different types of entities competed and there was no obvious way to
2413 settle it. Different controllers did different things.
2415 The cpu controller considered threads and cgroups as equivalents and
2416 mapped nice levels to cgroup weights. This worked for some cases but
2417 fell flat when children wanted to be allocated specific ratios of CPU
2418 cycles and the number of internal threads fluctuated - the ratios
2419 constantly changed as the number of competing entities fluctuated.
2420 There also were other issues. The mapping from nice level to weight
2421 wasn't obvious or universal, and there were various other knobs which
2422 simply weren't available for threads.
2424 The io controller implicitly created a hidden leaf node for each
2425 cgroup to host the threads. The hidden leaf had its own copies of all
2426 the knobs with ``leaf_`` prefixed. While this allowed equivalent
2427 control over internal threads, it was with serious drawbacks. It
2428 always added an extra layer of nesting which wouldn't be necessary
2429 otherwise, made the interface messy and significantly complicated the
2432 The memory controller didn't have a way to control what happened
2433 between internal tasks and child cgroups and the behavior was not
2434 clearly defined. There were attempts to add ad-hoc behaviors and
2435 knobs to tailor the behavior to specific workloads which would have
2436 led to problems extremely difficult to resolve in the long term.
2438 Multiple controllers struggled with internal tasks and came up with
2439 different ways to deal with it; unfortunately, all the approaches were
2440 severely flawed and, furthermore, the widely different behaviors
2441 made cgroup as a whole highly inconsistent.
2443 This clearly is a problem which needs to be addressed from cgroup core
2447 Other Interface Issues
2448 ----------------------
2450 cgroup v1 grew without oversight and developed a large number of
2451 idiosyncrasies and inconsistencies. One issue on the cgroup core side
2452 was how an empty cgroup was notified - a userland helper binary was
2453 forked and executed for each event. The event delivery wasn't
2454 recursive or delegatable. The limitations of the mechanism also led
2455 to in-kernel event delivery filtering mechanism further complicating
2458 Controller interfaces were problematic too. An extreme example is
2459 controllers completely ignoring hierarchical organization and treating
2460 all cgroups as if they were all located directly under the root
2461 cgroup. Some controllers exposed a large amount of inconsistent
2462 implementation details to userland.
2464 There also was no consistency across controllers. When a new cgroup
2465 was created, some controllers defaulted to not imposing extra
2466 restrictions while others disallowed any resource usage until
2467 explicitly configured. Configuration knobs for the same type of
2468 control used widely differing naming schemes and formats. Statistics
2469 and information knobs were named arbitrarily and used different
2470 formats and units even in the same controller.
2472 cgroup v2 establishes common conventions where appropriate and updates
2473 controllers so that they expose minimal and consistent interfaces.
2476 Controller Issues and Remedies
2477 ------------------------------
2482 The original lower boundary, the soft limit, is defined as a limit
2483 that is per default unset. As a result, the set of cgroups that
2484 global reclaim prefers is opt-in, rather than opt-out. The costs for
2485 optimizing these mostly negative lookups are so high that the
2486 implementation, despite its enormous size, does not even provide the
2487 basic desirable behavior. First off, the soft limit has no
2488 hierarchical meaning. All configured groups are organized in a global
2489 rbtree and treated like equal peers, regardless where they are located
2490 in the hierarchy. This makes subtree delegation impossible. Second,
2491 the soft limit reclaim pass is so aggressive that it not just
2492 introduces high allocation latencies into the system, but also impacts
2493 system performance due to overreclaim, to the point where the feature
2494 becomes self-defeating.
2496 The memory.low boundary on the other hand is a top-down allocated
2497 reserve. A cgroup enjoys reclaim protection when it's within its
2498 effective low, which makes delegation of subtrees possible. It also
2499 enjoys having reclaim pressure proportional to its overage when
2500 above its effective low.
2502 The original high boundary, the hard limit, is defined as a strict
2503 limit that can not budge, even if the OOM killer has to be called.
2504 But this generally goes against the goal of making the most out of the
2505 available memory. The memory consumption of workloads varies during
2506 runtime, and that requires users to overcommit. But doing that with a
2507 strict upper limit requires either a fairly accurate prediction of the
2508 working set size or adding slack to the limit. Since working set size
2509 estimation is hard and error prone, and getting it wrong results in
2510 OOM kills, most users tend to err on the side of a looser limit and
2511 end up wasting precious resources.
2513 The memory.high boundary on the other hand can be set much more
2514 conservatively. When hit, it throttles allocations by forcing them
2515 into direct reclaim to work off the excess, but it never invokes the
2516 OOM killer. As a result, a high boundary that is chosen too
2517 aggressively will not terminate the processes, but instead it will
2518 lead to gradual performance degradation. The user can monitor this
2519 and make corrections until the minimal memory footprint that still
2520 gives acceptable performance is found.
2522 In extreme cases, with many concurrent allocations and a complete
2523 breakdown of reclaim progress within the group, the high boundary can
2524 be exceeded. But even then it's mostly better to satisfy the
2525 allocation from the slack available in other groups or the rest of the
2526 system than killing the group. Otherwise, memory.max is there to
2527 limit this type of spillover and ultimately contain buggy or even
2528 malicious applications.
2530 Setting the original memory.limit_in_bytes below the current usage was
2531 subject to a race condition, where concurrent charges could cause the
2532 limit setting to fail. memory.max on the other hand will first set the
2533 limit to prevent new charges, and then reclaim and OOM kill until the
2534 new limit is met - or the task writing to memory.max is killed.
2536 The combined memory+swap accounting and limiting is replaced by real
2537 control over swap space.
2539 The main argument for a combined memory+swap facility in the original
2540 cgroup design was that global or parental pressure would always be
2541 able to swap all anonymous memory of a child group, regardless of the
2542 child's own (possibly untrusted) configuration. However, untrusted
2543 groups can sabotage swapping by other means - such as referencing its
2544 anonymous memory in a tight loop - and an admin can not assume full
2545 swappability when overcommitting untrusted jobs.
2547 For trusted jobs, on the other hand, a combined counter is not an
2548 intuitive userspace interface, and it flies in the face of the idea
2549 that cgroup controllers should account and limit specific physical
2550 resources. Swap space is a resource like all others in the system,
2551 and that's why unified hierarchy allows distributing it separately.