mm: treat memory.low value inclusive
[linux-block.git] / Documentation / admin-guide / cgroup-v2.rst
CommitLineData
633b11be 1================
6c292092 2Control Group v2
633b11be 3================
6c292092 4
633b11be
MCC
5:Date: October, 2015
6:Author: Tejun Heo <tj@kernel.org>
6c292092
TH
7
8This is the authoritative documentation on the design, interface and
9conventions of cgroup v2. It describes all userland-visible aspects
10of cgroup including core and specific controller behaviors. All
11future changes must be reflected in this document. Documentation for
9a2ddda5 12v1 is available under Documentation/cgroup-v1/.
6c292092 13
633b11be
MCC
14.. CONTENTS
15
16 1. Introduction
17 1-1. Terminology
18 1-2. What is cgroup?
19 2. Basic Operations
20 2-1. Mounting
8cfd8147
TH
21 2-2. Organizing Processes and Threads
22 2-2-1. Processes
23 2-2-2. Threads
633b11be
MCC
24 2-3. [Un]populated Notification
25 2-4. Controlling Controllers
26 2-4-1. Enabling and Disabling
27 2-4-2. Top-down Constraint
28 2-4-3. No Internal Process Constraint
29 2-5. Delegation
30 2-5-1. Model of Delegation
31 2-5-2. Delegation Containment
32 2-6. Guidelines
33 2-6-1. Organize Once and Control
34 2-6-2. Avoid Name Collisions
35 3. Resource Distribution Models
36 3-1. Weights
37 3-2. Limits
38 3-3. Protections
39 3-4. Allocations
40 4. Interface Files
41 4-1. Format
42 4-2. Conventions
43 4-3. Core Interface Files
44 5. Controllers
45 5-1. CPU
46 5-1-1. CPU Interface Files
47 5-2. Memory
48 5-2-1. Memory Interface Files
49 5-2-2. Usage Guidelines
50 5-2-3. Memory Ownership
51 5-3. IO
52 5-3-1. IO Interface Files
53 5-3-2. Writeback
54 5-4. PID
55 5-4-1. PID Interface Files
4ad5a321
RG
56 5-5. Device
57 5-6. RDMA
58 5-6-1. RDMA Interface Files
59 5-7. Misc
60 5-7-1. perf_event
c4e0842b
MS
61 5-N. Non-normative information
62 5-N-1. CPU controller root cgroup process behaviour
63 5-N-2. IO controller root cgroup process behaviour
633b11be
MCC
64 6. Namespace
65 6-1. Basics
66 6-2. The Root and Views
67 6-3. Migration and setns(2)
68 6-4. Interaction with Other Namespaces
69 P. Information on Kernel Programming
70 P-1. Filesystem Support for Writeback
71 D. Deprecated v1 Core Features
72 R. Issues with v1 and Rationales for v2
73 R-1. Multiple Hierarchies
74 R-2. Thread Granularity
75 R-3. Competition Between Inner Nodes and Threads
76 R-4. Other Interface Issues
77 R-5. Controller Issues and Remedies
78 R-5-1. Memory
79
80
81Introduction
82============
83
84Terminology
85-----------
6c292092
TH
86
87"cgroup" stands for "control group" and is never capitalized. The
88singular form is used to designate the whole feature and also as a
89qualifier as in "cgroup controllers". When explicitly referring to
90multiple individual control groups, the plural form "cgroups" is used.
91
92
633b11be
MCC
93What is cgroup?
94---------------
6c292092
TH
95
96cgroup is a mechanism to organize processes hierarchically and
97distribute system resources along the hierarchy in a controlled and
98configurable manner.
99
100cgroup is largely composed of two parts - the core and controllers.
101cgroup core is primarily responsible for hierarchically organizing
102processes. A cgroup controller is usually responsible for
103distributing a specific type of system resource along the hierarchy
104although there are utility controllers which serve purposes other than
105resource distribution.
106
107cgroups form a tree structure and every process in the system belongs
108to one and only one cgroup. All threads of a process belong to the
109same cgroup. On creation, all processes are put in the cgroup that
110the parent process belongs to at the time. A process can be migrated
111to another cgroup. Migration of a process doesn't affect already
112existing descendant processes.
113
114Following certain structural constraints, controllers may be enabled or
115disabled selectively on a cgroup. All controller behaviors are
116hierarchical - if a controller is enabled on a cgroup, it affects all
117processes which belong to the cgroups consisting the inclusive
118sub-hierarchy of the cgroup. When a controller is enabled on a nested
119cgroup, it always restricts the resource distribution further. The
120restrictions set closer to the root in the hierarchy can not be
121overridden from further away.
122
123
633b11be
MCC
124Basic Operations
125================
6c292092 126
633b11be
MCC
127Mounting
128--------
6c292092
TH
129
130Unlike v1, cgroup v2 has only single hierarchy. The cgroup v2
633b11be 131hierarchy can be mounted with the following mount command::
6c292092
TH
132
133 # mount -t cgroup2 none $MOUNT_POINT
134
135cgroup2 filesystem has the magic number 0x63677270 ("cgrp"). All
136controllers which support v2 and are not bound to a v1 hierarchy are
137automatically bound to the v2 hierarchy and show up at the root.
138Controllers which are not in active use in the v2 hierarchy can be
139bound to other hierarchies. This allows mixing v2 hierarchy with the
140legacy v1 multiple hierarchies in a fully backward compatible way.
141
142A controller can be moved across hierarchies only after the controller
143is no longer referenced in its current hierarchy. Because per-cgroup
144controller states are destroyed asynchronously and controllers may
145have lingering references, a controller may not show up immediately on
146the v2 hierarchy after the final umount of the previous hierarchy.
147Similarly, a controller should be fully disabled to be moved out of
148the unified hierarchy and it may take some time for the disabled
149controller to become available for other hierarchies; furthermore, due
150to inter-controller dependencies, other controllers may need to be
151disabled too.
152
153While useful for development and manual configurations, moving
154controllers dynamically between the v2 and other hierarchies is
155strongly discouraged for production use. It is recommended to decide
156the hierarchies and controller associations before starting using the
157controllers after system boot.
158
1619b6d4
JW
159During transition to v2, system management software might still
160automount the v1 cgroup filesystem and so hijack all controllers
161during boot, before manual intervention is possible. To make testing
162and experimenting easier, the kernel parameter cgroup_no_v1= allows
163disabling controllers in v1 and make them always available in v2.
164
5136f636
TH
165cgroup v2 currently supports the following mount options.
166
167 nsdelegate
168
169 Consider cgroup namespaces as delegation boundaries. This
170 option is system wide and can only be set on mount or modified
171 through remount from the init namespace. The mount option is
172 ignored on non-init namespace mounts. Please refer to the
173 Delegation section for details.
174
6c292092 175
8cfd8147
TH
176Organizing Processes and Threads
177--------------------------------
178
179Processes
180~~~~~~~~~
6c292092
TH
181
182Initially, only the root cgroup exists to which all processes belong.
633b11be 183A child cgroup can be created by creating a sub-directory::
6c292092
TH
184
185 # mkdir $CGROUP_NAME
186
187A given cgroup may have multiple child cgroups forming a tree
188structure. Each cgroup has a read-writable interface file
189"cgroup.procs". When read, it lists the PIDs of all processes which
190belong to the cgroup one-per-line. The PIDs are not ordered and the
191same PID may show up more than once if the process got moved to
192another cgroup and then back or the PID got recycled while reading.
193
194A process can be migrated into a cgroup by writing its PID to the
195target cgroup's "cgroup.procs" file. Only one process can be migrated
196on a single write(2) call. If a process is composed of multiple
197threads, writing the PID of any thread migrates all threads of the
198process.
199
200When a process forks a child process, the new process is born into the
201cgroup that the forking process belongs to at the time of the
202operation. After exit, a process stays associated with the cgroup
203that it belonged to at the time of exit until it's reaped; however, a
204zombie process does not appear in "cgroup.procs" and thus can't be
205moved to another cgroup.
206
207A cgroup which doesn't have any children or live processes can be
208destroyed by removing the directory. Note that a cgroup which doesn't
209have any children and is associated only with zombie processes is
633b11be 210considered empty and can be removed::
6c292092
TH
211
212 # rmdir $CGROUP_NAME
213
214"/proc/$PID/cgroup" lists a process's cgroup membership. If legacy
215cgroup is in use in the system, this file may contain multiple lines,
216one for each hierarchy. The entry for cgroup v2 is always in the
633b11be 217format "0::$PATH"::
6c292092
TH
218
219 # cat /proc/842/cgroup
220 ...
221 0::/test-cgroup/test-cgroup-nested
222
223If the process becomes a zombie and the cgroup it was associated with
633b11be 224is removed subsequently, " (deleted)" is appended to the path::
6c292092
TH
225
226 # cat /proc/842/cgroup
227 ...
228 0::/test-cgroup/test-cgroup-nested (deleted)
229
230
8cfd8147
TH
231Threads
232~~~~~~~
233
234cgroup v2 supports thread granularity for a subset of controllers to
235support use cases requiring hierarchical resource distribution across
236the threads of a group of processes. By default, all threads of a
237process belong to the same cgroup, which also serves as the resource
238domain to host resource consumptions which are not specific to a
239process or thread. The thread mode allows threads to be spread across
240a subtree while still maintaining the common resource domain for them.
241
242Controllers which support thread mode are called threaded controllers.
243The ones which don't are called domain controllers.
244
245Marking a cgroup threaded makes it join the resource domain of its
246parent as a threaded cgroup. The parent may be another threaded
247cgroup whose resource domain is further up in the hierarchy. The root
248of a threaded subtree, that is, the nearest ancestor which is not
249threaded, is called threaded domain or thread root interchangeably and
250serves as the resource domain for the entire subtree.
251
252Inside a threaded subtree, threads of a process can be put in
253different cgroups and are not subject to the no internal process
254constraint - threaded controllers can be enabled on non-leaf cgroups
255whether they have threads in them or not.
256
257As the threaded domain cgroup hosts all the domain resource
258consumptions of the subtree, it is considered to have internal
259resource consumptions whether there are processes in it or not and
260can't have populated child cgroups which aren't threaded. Because the
261root cgroup is not subject to no internal process constraint, it can
262serve both as a threaded domain and a parent to domain cgroups.
263
264The current operation mode or type of the cgroup is shown in the
265"cgroup.type" file which indicates whether the cgroup is a normal
266domain, a domain which is serving as the domain of a threaded subtree,
267or a threaded cgroup.
268
269On creation, a cgroup is always a domain cgroup and can be made
270threaded by writing "threaded" to the "cgroup.type" file. The
271operation is single direction::
272
273 # echo threaded > cgroup.type
274
275Once threaded, the cgroup can't be made a domain again. To enable the
276thread mode, the following conditions must be met.
277
278- As the cgroup will join the parent's resource domain. The parent
279 must either be a valid (threaded) domain or a threaded cgroup.
280
918a8c2c
TH
281- When the parent is an unthreaded domain, it must not have any domain
282 controllers enabled or populated domain children. The root is
283 exempt from this requirement.
8cfd8147
TH
284
285Topology-wise, a cgroup can be in an invalid state. Please consider
2877cbe6 286the following topology::
8cfd8147
TH
287
288 A (threaded domain) - B (threaded) - C (domain, just created)
289
290C is created as a domain but isn't connected to a parent which can
291host child domains. C can't be used until it is turned into a
292threaded cgroup. "cgroup.type" file will report "domain (invalid)" in
293these cases. Operations which fail due to invalid topology use
294EOPNOTSUPP as the errno.
295
296A domain cgroup is turned into a threaded domain when one of its child
297cgroup becomes threaded or threaded controllers are enabled in the
298"cgroup.subtree_control" file while there are processes in the cgroup.
299A threaded domain reverts to a normal domain when the conditions
300clear.
301
302When read, "cgroup.threads" contains the list of the thread IDs of all
303threads in the cgroup. Except that the operations are per-thread
304instead of per-process, "cgroup.threads" has the same format and
305behaves the same way as "cgroup.procs". While "cgroup.threads" can be
306written to in any cgroup, as it can only move threads inside the same
307threaded domain, its operations are confined inside each threaded
308subtree.
309
310The threaded domain cgroup serves as the resource domain for the whole
311subtree, and, while the threads can be scattered across the subtree,
312all the processes are considered to be in the threaded domain cgroup.
313"cgroup.procs" in a threaded domain cgroup contains the PIDs of all
314processes in the subtree and is not readable in the subtree proper.
315However, "cgroup.procs" can be written to from anywhere in the subtree
316to migrate all threads of the matching process to the cgroup.
317
318Only threaded controllers can be enabled in a threaded subtree. When
319a threaded controller is enabled inside a threaded subtree, it only
320accounts for and controls resource consumptions associated with the
321threads in the cgroup and its descendants. All consumptions which
322aren't tied to a specific thread belong to the threaded domain cgroup.
323
324Because a threaded subtree is exempt from no internal process
325constraint, a threaded controller must be able to handle competition
326between threads in a non-leaf cgroup and its child cgroups. Each
327threaded controller defines how such competitions are handled.
328
329
633b11be
MCC
330[Un]populated Notification
331--------------------------
6c292092
TH
332
333Each non-root cgroup has a "cgroup.events" file which contains
334"populated" field indicating whether the cgroup's sub-hierarchy has
335live processes in it. Its value is 0 if there is no live process in
336the cgroup and its descendants; otherwise, 1. poll and [id]notify
337events are triggered when the value changes. This can be used, for
338example, to start a clean-up operation after all processes of a given
339sub-hierarchy have exited. The populated state updates and
340notifications are recursive. Consider the following sub-hierarchy
341where the numbers in the parentheses represent the numbers of processes
633b11be 342in each cgroup::
6c292092
TH
343
344 A(4) - B(0) - C(1)
345 \ D(0)
346
347A, B and C's "populated" fields would be 1 while D's 0. After the one
348process in C exits, B and C's "populated" fields would flip to "0" and
349file modified events will be generated on the "cgroup.events" files of
350both cgroups.
351
352
633b11be
MCC
353Controlling Controllers
354-----------------------
6c292092 355
633b11be
MCC
356Enabling and Disabling
357~~~~~~~~~~~~~~~~~~~~~~
6c292092
TH
358
359Each cgroup has a "cgroup.controllers" file which lists all
633b11be 360controllers available for the cgroup to enable::
6c292092
TH
361
362 # cat cgroup.controllers
363 cpu io memory
364
365No controller is enabled by default. Controllers can be enabled and
633b11be 366disabled by writing to the "cgroup.subtree_control" file::
6c292092
TH
367
368 # echo "+cpu +memory -io" > cgroup.subtree_control
369
370Only controllers which are listed in "cgroup.controllers" can be
371enabled. When multiple operations are specified as above, either they
372all succeed or fail. If multiple operations on the same controller
373are specified, the last one is effective.
374
375Enabling a controller in a cgroup indicates that the distribution of
376the target resource across its immediate children will be controlled.
377Consider the following sub-hierarchy. The enabled controllers are
633b11be 378listed in parentheses::
6c292092
TH
379
380 A(cpu,memory) - B(memory) - C()
381 \ D()
382
383As A has "cpu" and "memory" enabled, A will control the distribution
384of CPU cycles and memory to its children, in this case, B. As B has
385"memory" enabled but not "CPU", C and D will compete freely on CPU
386cycles but their division of memory available to B will be controlled.
387
388As a controller regulates the distribution of the target resource to
389the cgroup's children, enabling it creates the controller's interface
390files in the child cgroups. In the above example, enabling "cpu" on B
391would create the "cpu." prefixed controller interface files in C and
392D. Likewise, disabling "memory" from B would remove the "memory."
393prefixed controller interface files from C and D. This means that the
394controller interface files - anything which doesn't start with
395"cgroup." are owned by the parent rather than the cgroup itself.
396
397
633b11be
MCC
398Top-down Constraint
399~~~~~~~~~~~~~~~~~~~
6c292092
TH
400
401Resources are distributed top-down and a cgroup can further distribute
402a resource only if the resource has been distributed to it from the
403parent. This means that all non-root "cgroup.subtree_control" files
404can only contain controllers which are enabled in the parent's
405"cgroup.subtree_control" file. A controller can be enabled only if
406the parent has the controller enabled and a controller can't be
407disabled if one or more children have it enabled.
408
409
633b11be
MCC
410No Internal Process Constraint
411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
6c292092 412
8cfd8147
TH
413Non-root cgroups can distribute domain resources to their children
414only when they don't have any processes of their own. In other words,
415only domain cgroups which don't contain any processes can have domain
416controllers enabled in their "cgroup.subtree_control" files.
6c292092 417
8cfd8147
TH
418This guarantees that, when a domain controller is looking at the part
419of the hierarchy which has it enabled, processes are always only on
420the leaves. This rules out situations where child cgroups compete
421against internal processes of the parent.
6c292092
TH
422
423The root cgroup is exempt from this restriction. Root contains
424processes and anonymous resource consumption which can't be associated
425with any other cgroups and requires special treatment from most
426controllers. How resource consumption in the root cgroup is governed
c4e0842b
MS
427is up to each controller (for more information on this topic please
428refer to the Non-normative information section in the Controllers
429chapter).
6c292092
TH
430
431Note that the restriction doesn't get in the way if there is no
432enabled controller in the cgroup's "cgroup.subtree_control". This is
433important as otherwise it wouldn't be possible to create children of a
434populated cgroup. To control resource distribution of a cgroup, the
435cgroup must create children and transfer all its processes to the
436children before enabling controllers in its "cgroup.subtree_control"
437file.
438
439
633b11be
MCC
440Delegation
441----------
6c292092 442
633b11be
MCC
443Model of Delegation
444~~~~~~~~~~~~~~~~~~~
6c292092 445
5136f636 446A cgroup can be delegated in two ways. First, to a less privileged
8cfd8147
TH
447user by granting write access of the directory and its "cgroup.procs",
448"cgroup.threads" and "cgroup.subtree_control" files to the user.
449Second, if the "nsdelegate" mount option is set, automatically to a
450cgroup namespace on namespace creation.
5136f636
TH
451
452Because the resource control interface files in a given directory
453control the distribution of the parent's resources, the delegatee
454shouldn't be allowed to write to them. For the first method, this is
455achieved by not granting access to these files. For the second, the
456kernel rejects writes to all files other than "cgroup.procs" and
457"cgroup.subtree_control" on a namespace root from inside the
458namespace.
459
460The end results are equivalent for both delegation types. Once
461delegated, the user can build sub-hierarchy under the directory,
462organize processes inside it as it sees fit and further distribute the
463resources it received from the parent. The limits and other settings
464of all resource controllers are hierarchical and regardless of what
465happens in the delegated sub-hierarchy, nothing can escape the
466resource restrictions imposed by the parent.
6c292092
TH
467
468Currently, cgroup doesn't impose any restrictions on the number of
469cgroups in or nesting depth of a delegated sub-hierarchy; however,
470this may be limited explicitly in the future.
471
472
633b11be
MCC
473Delegation Containment
474~~~~~~~~~~~~~~~~~~~~~~
6c292092
TH
475
476A delegated sub-hierarchy is contained in the sense that processes
5136f636
TH
477can't be moved into or out of the sub-hierarchy by the delegatee.
478
479For delegations to a less privileged user, this is achieved by
480requiring the following conditions for a process with a non-root euid
481to migrate a target process into a cgroup by writing its PID to the
482"cgroup.procs" file.
6c292092 483
6c292092
TH
484- The writer must have write access to the "cgroup.procs" file.
485
486- The writer must have write access to the "cgroup.procs" file of the
487 common ancestor of the source and destination cgroups.
488
576dd464 489The above two constraints ensure that while a delegatee may migrate
6c292092
TH
490processes around freely in the delegated sub-hierarchy it can't pull
491in from or push out to outside the sub-hierarchy.
492
493For an example, let's assume cgroups C0 and C1 have been delegated to
494user U0 who created C00, C01 under C0 and C10 under C1 as follows and
633b11be 495all processes under C0 and C1 belong to U0::
6c292092
TH
496
497 ~~~~~~~~~~~~~ - C0 - C00
498 ~ cgroup ~ \ C01
499 ~ hierarchy ~
500 ~~~~~~~~~~~~~ - C1 - C10
501
502Let's also say U0 wants to write the PID of a process which is
503currently in C10 into "C00/cgroup.procs". U0 has write access to the
576dd464
TH
504file; however, the common ancestor of the source cgroup C10 and the
505destination cgroup C00 is above the points of delegation and U0 would
506not have write access to its "cgroup.procs" files and thus the write
507will be denied with -EACCES.
6c292092 508
5136f636
TH
509For delegations to namespaces, containment is achieved by requiring
510that both the source and destination cgroups are reachable from the
511namespace of the process which is attempting the migration. If either
512is not reachable, the migration is rejected with -ENOENT.
513
6c292092 514
633b11be
MCC
515Guidelines
516----------
6c292092 517
633b11be
MCC
518Organize Once and Control
519~~~~~~~~~~~~~~~~~~~~~~~~~
6c292092
TH
520
521Migrating a process across cgroups is a relatively expensive operation
522and stateful resources such as memory are not moved together with the
523process. This is an explicit design decision as there often exist
524inherent trade-offs between migration and various hot paths in terms
525of synchronization cost.
526
527As such, migrating processes across cgroups frequently as a means to
528apply different resource restrictions is discouraged. A workload
529should be assigned to a cgroup according to the system's logical and
530resource structure once on start-up. Dynamic adjustments to resource
531distribution can be made by changing controller configuration through
532the interface files.
533
534
633b11be
MCC
535Avoid Name Collisions
536~~~~~~~~~~~~~~~~~~~~~
6c292092
TH
537
538Interface files for a cgroup and its children cgroups occupy the same
539directory and it is possible to create children cgroups which collide
540with interface files.
541
542All cgroup core interface files are prefixed with "cgroup." and each
543controller's interface files are prefixed with the controller name and
544a dot. A controller's name is composed of lower case alphabets and
545'_'s but never begins with an '_' so it can be used as the prefix
546character for collision avoidance. Also, interface file names won't
547start or end with terms which are often used in categorizing workloads
548such as job, service, slice, unit or workload.
549
550cgroup doesn't do anything to prevent name collisions and it's the
551user's responsibility to avoid them.
552
553
633b11be
MCC
554Resource Distribution Models
555============================
6c292092
TH
556
557cgroup controllers implement several resource distribution schemes
558depending on the resource type and expected use cases. This section
559describes major schemes in use along with their expected behaviors.
560
561
633b11be
MCC
562Weights
563-------
6c292092
TH
564
565A parent's resource is distributed by adding up the weights of all
566active children and giving each the fraction matching the ratio of its
567weight against the sum. As only children which can make use of the
568resource at the moment participate in the distribution, this is
569work-conserving. Due to the dynamic nature, this model is usually
570used for stateless resources.
571
572All weights are in the range [1, 10000] with the default at 100. This
573allows symmetric multiplicative biases in both directions at fine
574enough granularity while staying in the intuitive range.
575
576As long as the weight is in range, all configuration combinations are
577valid and there is no reason to reject configuration changes or
578process migrations.
579
580"cpu.weight" proportionally distributes CPU cycles to active children
581and is an example of this type.
582
583
633b11be
MCC
584Limits
585------
6c292092
TH
586
587A child can only consume upto the configured amount of the resource.
588Limits can be over-committed - the sum of the limits of children can
589exceed the amount of resource available to the parent.
590
591Limits are in the range [0, max] and defaults to "max", which is noop.
592
593As limits can be over-committed, all configuration combinations are
594valid and there is no reason to reject configuration changes or
595process migrations.
596
597"io.max" limits the maximum BPS and/or IOPS that a cgroup can consume
598on an IO device and is an example of this type.
599
600
633b11be
MCC
601Protections
602-----------
6c292092
TH
603
604A cgroup is protected to be allocated upto the configured amount of
605the resource if the usages of all its ancestors are under their
606protected levels. Protections can be hard guarantees or best effort
607soft boundaries. Protections can also be over-committed in which case
608only upto the amount available to the parent is protected among
609children.
610
611Protections are in the range [0, max] and defaults to 0, which is
612noop.
613
614As protections can be over-committed, all configuration combinations
615are valid and there is no reason to reject configuration changes or
616process migrations.
617
618"memory.low" implements best-effort memory protection and is an
619example of this type.
620
621
633b11be
MCC
622Allocations
623-----------
6c292092
TH
624
625A cgroup is exclusively allocated a certain amount of a finite
626resource. Allocations can't be over-committed - the sum of the
627allocations of children can not exceed the amount of resource
628available to the parent.
629
630Allocations are in the range [0, max] and defaults to 0, which is no
631resource.
632
633As allocations can't be over-committed, some configuration
634combinations are invalid and should be rejected. Also, if the
635resource is mandatory for execution of processes, process migrations
636may be rejected.
637
638"cpu.rt.max" hard-allocates realtime slices and is an example of this
639type.
640
641
633b11be
MCC
642Interface Files
643===============
6c292092 644
633b11be
MCC
645Format
646------
6c292092
TH
647
648All interface files should be in one of the following formats whenever
633b11be 649possible::
6c292092
TH
650
651 New-line separated values
652 (when only one value can be written at once)
653
654 VAL0\n
655 VAL1\n
656 ...
657
658 Space separated values
659 (when read-only or multiple values can be written at once)
660
661 VAL0 VAL1 ...\n
662
663 Flat keyed
664
665 KEY0 VAL0\n
666 KEY1 VAL1\n
667 ...
668
669 Nested keyed
670
671 KEY0 SUB_KEY0=VAL00 SUB_KEY1=VAL01...
672 KEY1 SUB_KEY0=VAL10 SUB_KEY1=VAL11...
673 ...
674
675For a writable file, the format for writing should generally match
676reading; however, controllers may allow omitting later fields or
677implement restricted shortcuts for most common use cases.
678
679For both flat and nested keyed files, only the values for a single key
680can be written at a time. For nested keyed files, the sub key pairs
681may be specified in any order and not all pairs have to be specified.
682
683
633b11be
MCC
684Conventions
685-----------
6c292092
TH
686
687- Settings for a single feature should be contained in a single file.
688
689- The root cgroup should be exempt from resource control and thus
690 shouldn't have resource control interface files. Also,
691 informational files on the root cgroup which end up showing global
692 information available elsewhere shouldn't exist.
693
694- If a controller implements weight based resource distribution, its
695 interface file should be named "weight" and have the range [1,
696 10000] with 100 as the default. The values are chosen to allow
697 enough and symmetric bias in both directions while keeping it
698 intuitive (the default is 100%).
699
700- If a controller implements an absolute resource guarantee and/or
701 limit, the interface files should be named "min" and "max"
702 respectively. If a controller implements best effort resource
703 guarantee and/or limit, the interface files should be named "low"
704 and "high" respectively.
705
706 In the above four control files, the special token "max" should be
707 used to represent upward infinity for both reading and writing.
708
709- If a setting has a configurable default value and keyed specific
710 overrides, the default entry should be keyed with "default" and
711 appear as the first entry in the file.
712
713 The default value can be updated by writing either "default $VAL" or
714 "$VAL".
715
716 When writing to update a specific override, "default" can be used as
717 the value to indicate removal of the override. Override entries
718 with "default" as the value must not appear when read.
719
720 For example, a setting which is keyed by major:minor device numbers
633b11be 721 with integer values may look like the following::
6c292092
TH
722
723 # cat cgroup-example-interface-file
724 default 150
725 8:0 300
726
633b11be 727 The default value can be updated by::
6c292092
TH
728
729 # echo 125 > cgroup-example-interface-file
730
633b11be 731 or::
6c292092
TH
732
733 # echo "default 125" > cgroup-example-interface-file
734
633b11be 735 An override can be set by::
6c292092
TH
736
737 # echo "8:16 170" > cgroup-example-interface-file
738
633b11be 739 and cleared by::
6c292092
TH
740
741 # echo "8:0 default" > cgroup-example-interface-file
742 # cat cgroup-example-interface-file
743 default 125
744 8:16 170
745
746- For events which are not very high frequency, an interface file
747 "events" should be created which lists event key value pairs.
748 Whenever a notifiable event happens, file modified event should be
749 generated on the file.
750
751
633b11be
MCC
752Core Interface Files
753--------------------
6c292092
TH
754
755All cgroup core files are prefixed with "cgroup."
756
8cfd8147
TH
757 cgroup.type
758
759 A read-write single value file which exists on non-root
760 cgroups.
761
762 When read, it indicates the current type of the cgroup, which
763 can be one of the following values.
764
765 - "domain" : A normal valid domain cgroup.
766
767 - "domain threaded" : A threaded domain cgroup which is
768 serving as the root of a threaded subtree.
769
770 - "domain invalid" : A cgroup which is in an invalid state.
771 It can't be populated or have controllers enabled. It may
772 be allowed to become a threaded cgroup.
773
774 - "threaded" : A threaded cgroup which is a member of a
775 threaded subtree.
776
777 A cgroup can be turned into a threaded cgroup by writing
778 "threaded" to this file.
779
6c292092 780 cgroup.procs
6c292092
TH
781 A read-write new-line separated values file which exists on
782 all cgroups.
783
784 When read, it lists the PIDs of all processes which belong to
785 the cgroup one-per-line. The PIDs are not ordered and the
786 same PID may show up more than once if the process got moved
787 to another cgroup and then back or the PID got recycled while
788 reading.
789
790 A PID can be written to migrate the process associated with
791 the PID to the cgroup. The writer should match all of the
792 following conditions.
793
6c292092 794 - It must have write access to the "cgroup.procs" file.
8cfd8147
TH
795
796 - It must have write access to the "cgroup.procs" file of the
797 common ancestor of the source and destination cgroups.
798
799 When delegating a sub-hierarchy, write access to this file
800 should be granted along with the containing directory.
801
802 In a threaded cgroup, reading this file fails with EOPNOTSUPP
803 as all the processes belong to the thread root. Writing is
804 supported and moves every thread of the process to the cgroup.
805
806 cgroup.threads
807 A read-write new-line separated values file which exists on
808 all cgroups.
809
810 When read, it lists the TIDs of all threads which belong to
811 the cgroup one-per-line. The TIDs are not ordered and the
812 same TID may show up more than once if the thread got moved to
813 another cgroup and then back or the TID got recycled while
814 reading.
815
816 A TID can be written to migrate the thread associated with the
817 TID to the cgroup. The writer should match all of the
818 following conditions.
819
820 - It must have write access to the "cgroup.threads" file.
821
822 - The cgroup that the thread is currently in must be in the
823 same resource domain as the destination cgroup.
6c292092
TH
824
825 - It must have write access to the "cgroup.procs" file of the
826 common ancestor of the source and destination cgroups.
827
828 When delegating a sub-hierarchy, write access to this file
829 should be granted along with the containing directory.
830
831 cgroup.controllers
6c292092
TH
832 A read-only space separated values file which exists on all
833 cgroups.
834
835 It shows space separated list of all controllers available to
836 the cgroup. The controllers are not ordered.
837
838 cgroup.subtree_control
6c292092
TH
839 A read-write space separated values file which exists on all
840 cgroups. Starts out empty.
841
842 When read, it shows space separated list of the controllers
843 which are enabled to control resource distribution from the
844 cgroup to its children.
845
846 Space separated list of controllers prefixed with '+' or '-'
847 can be written to enable or disable controllers. A controller
848 name prefixed with '+' enables the controller and '-'
849 disables. If a controller appears more than once on the list,
850 the last one is effective. When multiple enable and disable
851 operations are specified, either all succeed or all fail.
852
853 cgroup.events
6c292092
TH
854 A read-only flat-keyed file which exists on non-root cgroups.
855 The following entries are defined. Unless specified
856 otherwise, a value change in this file generates a file
857 modified event.
858
859 populated
6c292092
TH
860 1 if the cgroup or its descendants contains any live
861 processes; otherwise, 0.
862
1a926e0b
RG
863 cgroup.max.descendants
864 A read-write single value files. The default is "max".
865
866 Maximum allowed number of descent cgroups.
867 If the actual number of descendants is equal or larger,
868 an attempt to create a new cgroup in the hierarchy will fail.
869
870 cgroup.max.depth
871 A read-write single value files. The default is "max".
872
873 Maximum allowed descent depth below the current cgroup.
874 If the actual descent depth is equal or larger,
875 an attempt to create a new child cgroup will fail.
876
ec39225c
RG
877 cgroup.stat
878 A read-only flat-keyed file with the following entries:
879
880 nr_descendants
881 Total number of visible descendant cgroups.
882
883 nr_dying_descendants
884 Total number of dying descendant cgroups. A cgroup becomes
885 dying after being deleted by a user. The cgroup will remain
886 in dying state for some time undefined time (which can depend
887 on system load) before being completely destroyed.
888
889 A process can't enter a dying cgroup under any circumstances,
890 a dying cgroup can't revive.
891
892 A dying cgroup can consume system resources not exceeding
893 limits, which were active at the moment of cgroup deletion.
894
6c292092 895
633b11be
MCC
896Controllers
897===========
6c292092 898
633b11be
MCC
899CPU
900---
6c292092 901
6c292092
TH
902The "cpu" controllers regulates distribution of CPU cycles. This
903controller implements weight and absolute bandwidth limit models for
904normal scheduling policy and absolute bandwidth allocation model for
905realtime scheduling policy.
906
c2f31b79
TH
907WARNING: cgroup2 doesn't yet support control of realtime processes and
908the cpu controller can only be enabled when all RT processes are in
909the root cgroup. Be aware that system management software may already
910have placed RT processes into nonroot cgroups during the system boot
911process, and these processes may need to be moved to the root cgroup
912before the cpu controller can be enabled.
913
6c292092 914
633b11be
MCC
915CPU Interface Files
916~~~~~~~~~~~~~~~~~~~
6c292092
TH
917
918All time durations are in microseconds.
919
920 cpu.stat
6c292092 921 A read-only flat-keyed file which exists on non-root cgroups.
d41bf8c9 922 This file exists whether the controller is enabled or not.
6c292092 923
d41bf8c9 924 It always reports the following three stats:
6c292092 925
633b11be
MCC
926 - usage_usec
927 - user_usec
928 - system_usec
d41bf8c9
TH
929
930 and the following three when the controller is enabled:
931
633b11be
MCC
932 - nr_periods
933 - nr_throttled
934 - throttled_usec
6c292092
TH
935
936 cpu.weight
6c292092
TH
937 A read-write single value file which exists on non-root
938 cgroups. The default is "100".
939
940 The weight in the range [1, 10000].
941
0d593634
TH
942 cpu.weight.nice
943 A read-write single value file which exists on non-root
944 cgroups. The default is "0".
945
946 The nice value is in the range [-20, 19].
947
948 This interface file is an alternative interface for
949 "cpu.weight" and allows reading and setting weight using the
950 same values used by nice(2). Because the range is smaller and
951 granularity is coarser for the nice values, the read value is
952 the closest approximation of the current weight.
953
6c292092 954 cpu.max
6c292092
TH
955 A read-write two value file which exists on non-root cgroups.
956 The default is "max 100000".
957
633b11be 958 The maximum bandwidth limit. It's in the following format::
6c292092
TH
959
960 $MAX $PERIOD
961
962 which indicates that the group may consume upto $MAX in each
963 $PERIOD duration. "max" for $MAX indicates no limit. If only
964 one number is written, $MAX is updated.
965
6c292092 966
633b11be
MCC
967Memory
968------
6c292092
TH
969
970The "memory" controller regulates distribution of memory. Memory is
971stateful and implements both limit and protection models. Due to the
972intertwining between memory usage and reclaim pressure and the
973stateful nature of memory, the distribution model is relatively
974complex.
975
976While not completely water-tight, all major memory usages by a given
977cgroup are tracked so that the total memory consumption can be
978accounted and controlled to a reasonable extent. Currently, the
979following types of memory usages are tracked.
980
981- Userland memory - page cache and anonymous memory.
982
983- Kernel data structures such as dentries and inodes.
984
985- TCP socket buffers.
986
987The above list may expand in the future for better coverage.
988
989
633b11be
MCC
990Memory Interface Files
991~~~~~~~~~~~~~~~~~~~~~~
6c292092
TH
992
993All memory amounts are in bytes. If a value which is not aligned to
994PAGE_SIZE is written, the value may be rounded up to the closest
995PAGE_SIZE multiple when read back.
996
997 memory.current
6c292092
TH
998 A read-only single value file which exists on non-root
999 cgroups.
1000
1001 The total amount of memory currently being used by the cgroup
1002 and its descendants.
1003
1004 memory.low
6c292092
TH
1005 A read-write single value file which exists on non-root
1006 cgroups. The default is "0".
1007
1008 Best-effort memory protection. If the memory usages of a
1009 cgroup and all its ancestors are below their low boundaries,
1010 the cgroup's memory won't be reclaimed unless memory can be
1011 reclaimed from unprotected cgroups.
1012
1013 Putting more memory than generally available under this
1014 protection is discouraged.
1015
1016 memory.high
6c292092
TH
1017 A read-write single value file which exists on non-root
1018 cgroups. The default is "max".
1019
1020 Memory usage throttle limit. This is the main mechanism to
1021 control memory usage of a cgroup. If a cgroup's usage goes
1022 over the high boundary, the processes of the cgroup are
1023 throttled and put under heavy reclaim pressure.
1024
1025 Going over the high limit never invokes the OOM killer and
1026 under extreme conditions the limit may be breached.
1027
1028 memory.max
6c292092
TH
1029 A read-write single value file which exists on non-root
1030 cgroups. The default is "max".
1031
1032 Memory usage hard limit. This is the final protection
1033 mechanism. If a cgroup's memory usage reaches this limit and
1034 can't be reduced, the OOM killer is invoked in the cgroup.
1035 Under certain circumstances, the usage may go over the limit
1036 temporarily.
1037
1038 This is the ultimate protection mechanism. As long as the
1039 high limit is used and monitored properly, this limit's
1040 utility is limited to providing the final safety net.
1041
1042 memory.events
6c292092
TH
1043 A read-only flat-keyed file which exists on non-root cgroups.
1044 The following entries are defined. Unless specified
1045 otherwise, a value change in this file generates a file
1046 modified event.
1047
1048 low
6c292092
TH
1049 The number of times the cgroup is reclaimed due to
1050 high memory pressure even though its usage is under
1051 the low boundary. This usually indicates that the low
1052 boundary is over-committed.
1053
1054 high
6c292092
TH
1055 The number of times processes of the cgroup are
1056 throttled and routed to perform direct memory reclaim
1057 because the high memory boundary was exceeded. For a
1058 cgroup whose memory usage is capped by the high limit
1059 rather than global memory pressure, this event's
1060 occurrences are expected.
1061
1062 max
6c292092
TH
1063 The number of times the cgroup's memory usage was
1064 about to go over the max boundary. If direct reclaim
8e675f7a 1065 fails to bring it down, the cgroup goes to OOM state.
6c292092
TH
1066
1067 oom
8e675f7a
KK
1068 The number of time the cgroup's memory usage was
1069 reached the limit and allocation was about to fail.
1070
1071 Depending on context result could be invocation of OOM
2877cbe6 1072 killer and retrying allocation or failing allocation.
8e675f7a
KK
1073
1074 Failed allocation in its turn could be returned into
2877cbe6 1075 userspace as -ENOMEM or silently ignored in cases like
633b11be 1076 disk readahead. For now OOM in memory cgroup kills
8e675f7a
KK
1077 tasks iff shortage has happened inside page fault.
1078
1079 oom_kill
8e675f7a
KK
1080 The number of processes belonging to this cgroup
1081 killed by any kind of OOM killer.
6c292092 1082
587d9f72 1083 memory.stat
587d9f72
JW
1084 A read-only flat-keyed file which exists on non-root cgroups.
1085
1086 This breaks down the cgroup's memory footprint into different
1087 types of memory, type-specific details, and other information
1088 on the state and past events of the memory management system.
1089
1090 All memory amounts are in bytes.
1091
1092 The entries are ordered to be human readable, and new entries
1093 can show up in the middle. Don't rely on items remaining in a
1094 fixed position; use the keys to look up specific values!
1095
1096 anon
587d9f72
JW
1097 Amount of memory used in anonymous mappings such as
1098 brk(), sbrk(), and mmap(MAP_ANONYMOUS)
1099
1100 file
587d9f72
JW
1101 Amount of memory used to cache filesystem data,
1102 including tmpfs and shared memory.
1103
12580e4b 1104 kernel_stack
12580e4b
VD
1105 Amount of memory allocated to kernel stacks.
1106
27ee57c9 1107 slab
27ee57c9
VD
1108 Amount of memory used for storing in-kernel data
1109 structures.
1110
4758e198 1111 sock
4758e198
JW
1112 Amount of memory used in network transmission buffers
1113
9a4caf1e 1114 shmem
9a4caf1e
JW
1115 Amount of cached filesystem data that is swap-backed,
1116 such as tmpfs, shm segments, shared anonymous mmap()s
1117
587d9f72 1118 file_mapped
587d9f72
JW
1119 Amount of cached filesystem data mapped with mmap()
1120
1121 file_dirty
587d9f72
JW
1122 Amount of cached filesystem data that was modified but
1123 not yet written back to disk
1124
1125 file_writeback
587d9f72
JW
1126 Amount of cached filesystem data that was modified and
1127 is currently being written back to disk
1128
633b11be 1129 inactive_anon, active_anon, inactive_file, active_file, unevictable
587d9f72
JW
1130 Amount of memory, swap-backed and filesystem-backed,
1131 on the internal memory management lists used by the
1132 page reclaim algorithm
1133
27ee57c9 1134 slab_reclaimable
27ee57c9
VD
1135 Part of "slab" that might be reclaimed, such as
1136 dentries and inodes.
1137
1138 slab_unreclaimable
27ee57c9
VD
1139 Part of "slab" that cannot be reclaimed on memory
1140 pressure.
1141
587d9f72 1142 pgfault
587d9f72
JW
1143 Total number of page faults incurred
1144
1145 pgmajfault
587d9f72
JW
1146 Number of major page faults incurred
1147
b340959e
RG
1148 workingset_refault
1149
1150 Number of refaults of previously evicted pages
1151
1152 workingset_activate
1153
1154 Number of refaulted pages that were immediately activated
1155
1156 workingset_nodereclaim
1157
1158 Number of times a shadow node has been reclaimed
1159
2262185c
RG
1160 pgrefill
1161
1162 Amount of scanned pages (in an active LRU list)
1163
1164 pgscan
1165
1166 Amount of scanned pages (in an inactive LRU list)
1167
1168 pgsteal
1169
1170 Amount of reclaimed pages
1171
1172 pgactivate
1173
1174 Amount of pages moved to the active LRU list
1175
1176 pgdeactivate
1177
1178 Amount of pages moved to the inactive LRU lis
1179
1180 pglazyfree
1181
1182 Amount of pages postponed to be freed under memory pressure
1183
1184 pglazyfreed
1185
1186 Amount of reclaimed lazyfree pages
1187
3e24b19d 1188 memory.swap.current
3e24b19d
VD
1189 A read-only single value file which exists on non-root
1190 cgroups.
1191
1192 The total amount of swap currently being used by the cgroup
1193 and its descendants.
1194
1195 memory.swap.max
3e24b19d
VD
1196 A read-write single value file which exists on non-root
1197 cgroups. The default is "max".
1198
1199 Swap usage hard limit. If a cgroup's swap usage reaches this
2877cbe6 1200 limit, anonymous memory of the cgroup will not be swapped out.
3e24b19d 1201
f3a53a3a
TH
1202 memory.swap.events
1203 A read-only flat-keyed file which exists on non-root cgroups.
1204 The following entries are defined. Unless specified
1205 otherwise, a value change in this file generates a file
1206 modified event.
1207
1208 max
1209 The number of times the cgroup's swap usage was about
1210 to go over the max boundary and swap allocation
1211 failed.
1212
1213 fail
1214 The number of times swap allocation failed either
1215 because of running out of swap system-wide or max
1216 limit.
1217
6c292092 1218
633b11be
MCC
1219Usage Guidelines
1220~~~~~~~~~~~~~~~~
6c292092
TH
1221
1222"memory.high" is the main mechanism to control memory usage.
1223Over-committing on high limit (sum of high limits > available memory)
1224and letting global memory pressure to distribute memory according to
1225usage is a viable strategy.
1226
1227Because breach of the high limit doesn't trigger the OOM killer but
1228throttles the offending cgroup, a management agent has ample
1229opportunities to monitor and take appropriate actions such as granting
1230more memory or terminating the workload.
1231
1232Determining whether a cgroup has enough memory is not trivial as
1233memory usage doesn't indicate whether the workload can benefit from
1234more memory. For example, a workload which writes data received from
1235network to a file can use all available memory but can also operate as
1236performant with a small amount of memory. A measure of memory
1237pressure - how much the workload is being impacted due to lack of
1238memory - is necessary to determine whether a workload needs more
1239memory; unfortunately, memory pressure monitoring mechanism isn't
1240implemented yet.
1241
1242
633b11be
MCC
1243Memory Ownership
1244~~~~~~~~~~~~~~~~
6c292092
TH
1245
1246A memory area is charged to the cgroup which instantiated it and stays
1247charged to the cgroup until the area is released. Migrating a process
1248to a different cgroup doesn't move the memory usages that it
1249instantiated while in the previous cgroup to the new cgroup.
1250
1251A memory area may be used by processes belonging to different cgroups.
1252To which cgroup the area will be charged is in-deterministic; however,
1253over time, the memory area is likely to end up in a cgroup which has
1254enough memory allowance to avoid high reclaim pressure.
1255
1256If a cgroup sweeps a considerable amount of memory which is expected
1257to be accessed repeatedly by other cgroups, it may make sense to use
1258POSIX_FADV_DONTNEED to relinquish the ownership of memory areas
1259belonging to the affected files to ensure correct memory ownership.
1260
1261
633b11be
MCC
1262IO
1263--
6c292092
TH
1264
1265The "io" controller regulates the distribution of IO resources. This
1266controller implements both weight based and absolute bandwidth or IOPS
1267limit distribution; however, weight based distribution is available
1268only if cfq-iosched is in use and neither scheme is available for
1269blk-mq devices.
1270
1271
633b11be
MCC
1272IO Interface Files
1273~~~~~~~~~~~~~~~~~~
6c292092
TH
1274
1275 io.stat
6c292092
TH
1276 A read-only nested-keyed file which exists on non-root
1277 cgroups.
1278
1279 Lines are keyed by $MAJ:$MIN device numbers and not ordered.
1280 The following nested keys are defined.
1281
633b11be 1282 ====== ===================
6c292092
TH
1283 rbytes Bytes read
1284 wbytes Bytes written
1285 rios Number of read IOs
1286 wios Number of write IOs
633b11be 1287 ====== ===================
6c292092 1288
633b11be 1289 An example read output follows:
6c292092
TH
1290
1291 8:16 rbytes=1459200 wbytes=314773504 rios=192 wios=353
1292 8:0 rbytes=90430464 wbytes=299008000 rios=8950 wios=1252
1293
1294 io.weight
6c292092
TH
1295 A read-write flat-keyed file which exists on non-root cgroups.
1296 The default is "default 100".
1297
1298 The first line is the default weight applied to devices
1299 without specific override. The rest are overrides keyed by
1300 $MAJ:$MIN device numbers and not ordered. The weights are in
1301 the range [1, 10000] and specifies the relative amount IO time
1302 the cgroup can use in relation to its siblings.
1303
1304 The default weight can be updated by writing either "default
1305 $WEIGHT" or simply "$WEIGHT". Overrides can be set by writing
1306 "$MAJ:$MIN $WEIGHT" and unset by writing "$MAJ:$MIN default".
1307
633b11be 1308 An example read output follows::
6c292092
TH
1309
1310 default 100
1311 8:16 200
1312 8:0 50
1313
1314 io.max
6c292092
TH
1315 A read-write nested-keyed file which exists on non-root
1316 cgroups.
1317
1318 BPS and IOPS based IO limit. Lines are keyed by $MAJ:$MIN
1319 device numbers and not ordered. The following nested keys are
1320 defined.
1321
633b11be 1322 ===== ==================================
6c292092
TH
1323 rbps Max read bytes per second
1324 wbps Max write bytes per second
1325 riops Max read IO operations per second
1326 wiops Max write IO operations per second
633b11be 1327 ===== ==================================
6c292092
TH
1328
1329 When writing, any number of nested key-value pairs can be
1330 specified in any order. "max" can be specified as the value
1331 to remove a specific limit. If the same key is specified
1332 multiple times, the outcome is undefined.
1333
1334 BPS and IOPS are measured in each IO direction and IOs are
1335 delayed if limit is reached. Temporary bursts are allowed.
1336
633b11be 1337 Setting read limit at 2M BPS and write at 120 IOPS for 8:16::
6c292092
TH
1338
1339 echo "8:16 rbps=2097152 wiops=120" > io.max
1340
633b11be 1341 Reading returns the following::
6c292092
TH
1342
1343 8:16 rbps=2097152 wbps=max riops=max wiops=120
1344
633b11be 1345 Write IOPS limit can be removed by writing the following::
6c292092
TH
1346
1347 echo "8:16 wiops=max" > io.max
1348
633b11be 1349 Reading now returns the following::
6c292092
TH
1350
1351 8:16 rbps=2097152 wbps=max riops=max wiops=max
1352
1353
633b11be
MCC
1354Writeback
1355~~~~~~~~~
6c292092
TH
1356
1357Page cache is dirtied through buffered writes and shared mmaps and
1358written asynchronously to the backing filesystem by the writeback
1359mechanism. Writeback sits between the memory and IO domains and
1360regulates the proportion of dirty memory by balancing dirtying and
1361write IOs.
1362
1363The io controller, in conjunction with the memory controller,
1364implements control of page cache writeback IOs. The memory controller
1365defines the memory domain that dirty memory ratio is calculated and
1366maintained for and the io controller defines the io domain which
1367writes out dirty pages for the memory domain. Both system-wide and
1368per-cgroup dirty memory states are examined and the more restrictive
1369of the two is enforced.
1370
1371cgroup writeback requires explicit support from the underlying
1372filesystem. Currently, cgroup writeback is implemented on ext2, ext4
1373and btrfs. On other filesystems, all writeback IOs are attributed to
1374the root cgroup.
1375
1376There are inherent differences in memory and writeback management
1377which affects how cgroup ownership is tracked. Memory is tracked per
1378page while writeback per inode. For the purpose of writeback, an
1379inode is assigned to a cgroup and all IO requests to write dirty pages
1380from the inode are attributed to that cgroup.
1381
1382As cgroup ownership for memory is tracked per page, there can be pages
1383which are associated with different cgroups than the one the inode is
1384associated with. These are called foreign pages. The writeback
1385constantly keeps track of foreign pages and, if a particular foreign
1386cgroup becomes the majority over a certain period of time, switches
1387the ownership of the inode to that cgroup.
1388
1389While this model is enough for most use cases where a given inode is
1390mostly dirtied by a single cgroup even when the main writing cgroup
1391changes over time, use cases where multiple cgroups write to a single
1392inode simultaneously are not supported well. In such circumstances, a
1393significant portion of IOs are likely to be attributed incorrectly.
1394As memory controller assigns page ownership on the first use and
1395doesn't update it until the page is released, even if writeback
1396strictly follows page ownership, multiple cgroups dirtying overlapping
1397areas wouldn't work as expected. It's recommended to avoid such usage
1398patterns.
1399
1400The sysctl knobs which affect writeback behavior are applied to cgroup
1401writeback as follows.
1402
633b11be 1403 vm.dirty_background_ratio, vm.dirty_ratio
6c292092
TH
1404 These ratios apply the same to cgroup writeback with the
1405 amount of available memory capped by limits imposed by the
1406 memory controller and system-wide clean memory.
1407
633b11be 1408 vm.dirty_background_bytes, vm.dirty_bytes
6c292092
TH
1409 For cgroup writeback, this is calculated into ratio against
1410 total available memory and applied the same way as
1411 vm.dirty[_background]_ratio.
1412
1413
633b11be
MCC
1414PID
1415---
20c56e59
HR
1416
1417The process number controller is used to allow a cgroup to stop any
1418new tasks from being fork()'d or clone()'d after a specified limit is
1419reached.
1420
1421The number of tasks in a cgroup can be exhausted in ways which other
1422controllers cannot prevent, thus warranting its own controller. For
1423example, a fork bomb is likely to exhaust the number of tasks before
1424hitting memory restrictions.
1425
1426Note that PIDs used in this controller refer to TIDs, process IDs as
1427used by the kernel.
1428
1429
633b11be
MCC
1430PID Interface Files
1431~~~~~~~~~~~~~~~~~~~
20c56e59
HR
1432
1433 pids.max
312eb712
TK
1434 A read-write single value file which exists on non-root
1435 cgroups. The default is "max".
20c56e59 1436
312eb712 1437 Hard limit of number of processes.
20c56e59
HR
1438
1439 pids.current
312eb712 1440 A read-only single value file which exists on all cgroups.
20c56e59 1441
312eb712
TK
1442 The number of processes currently in the cgroup and its
1443 descendants.
20c56e59
HR
1444
1445Organisational operations are not blocked by cgroup policies, so it is
1446possible to have pids.current > pids.max. This can be done by either
1447setting the limit to be smaller than pids.current, or attaching enough
1448processes to the cgroup such that pids.current is larger than
1449pids.max. However, it is not possible to violate a cgroup PID policy
1450through fork() or clone(). These will return -EAGAIN if the creation
1451of a new process would cause a cgroup policy to be violated.
1452
1453
4ad5a321
RG
1454Device controller
1455-----------------
1456
1457Device controller manages access to device files. It includes both
1458creation of new device files (using mknod), and access to the
1459existing device files.
1460
1461Cgroup v2 device controller has no interface files and is implemented
1462on top of cgroup BPF. To control access to device files, a user may
1463create bpf programs of the BPF_CGROUP_DEVICE type and attach them
1464to cgroups. On an attempt to access a device file, corresponding
1465BPF programs will be executed, and depending on the return value
1466the attempt will succeed or fail with -EPERM.
1467
1468A BPF_CGROUP_DEVICE program takes a pointer to the bpf_cgroup_dev_ctx
1469structure, which describes the device access attempt: access type
1470(mknod/read/write) and device (type, major and minor numbers).
1471If the program returns 0, the attempt fails with -EPERM, otherwise
1472it succeeds.
1473
1474An example of BPF_CGROUP_DEVICE program may be found in the kernel
1475source tree in the tools/testing/selftests/bpf/dev_cgroup.c file.
1476
1477
633b11be
MCC
1478RDMA
1479----
968ebff1 1480
9c1e67f9
PP
1481The "rdma" controller regulates the distribution and accounting of
1482of RDMA resources.
1483
633b11be
MCC
1484RDMA Interface Files
1485~~~~~~~~~~~~~~~~~~~~
9c1e67f9
PP
1486
1487 rdma.max
1488 A readwrite nested-keyed file that exists for all the cgroups
1489 except root that describes current configured resource limit
1490 for a RDMA/IB device.
1491
1492 Lines are keyed by device name and are not ordered.
1493 Each line contains space separated resource name and its configured
1494 limit that can be distributed.
1495
1496 The following nested keys are defined.
1497
633b11be 1498 ========== =============================
9c1e67f9
PP
1499 hca_handle Maximum number of HCA Handles
1500 hca_object Maximum number of HCA Objects
633b11be 1501 ========== =============================
9c1e67f9 1502
633b11be 1503 An example for mlx4 and ocrdma device follows::
9c1e67f9
PP
1504
1505 mlx4_0 hca_handle=2 hca_object=2000
1506 ocrdma1 hca_handle=3 hca_object=max
1507
1508 rdma.current
1509 A read-only file that describes current resource usage.
1510 It exists for all the cgroup except root.
1511
633b11be 1512 An example for mlx4 and ocrdma device follows::
9c1e67f9
PP
1513
1514 mlx4_0 hca_handle=1 hca_object=20
1515 ocrdma1 hca_handle=1 hca_object=23
1516
1517
633b11be
MCC
1518Misc
1519----
63f1ca59 1520
633b11be
MCC
1521perf_event
1522~~~~~~~~~~
968ebff1
TH
1523
1524perf_event controller, if not mounted on a legacy hierarchy, is
1525automatically enabled on the v2 hierarchy so that perf events can
1526always be filtered by cgroup v2 path. The controller can still be
1527moved to a legacy hierarchy after v2 hierarchy is populated.
1528
1529
c4e0842b
MS
1530Non-normative information
1531-------------------------
1532
1533This section contains information that isn't considered to be a part of
1534the stable kernel API and so is subject to change.
1535
1536
1537CPU controller root cgroup process behaviour
1538~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1539
1540When distributing CPU cycles in the root cgroup each thread in this
1541cgroup is treated as if it was hosted in a separate child cgroup of the
1542root cgroup. This child cgroup weight is dependent on its thread nice
1543level.
1544
1545For details of this mapping see sched_prio_to_weight array in
1546kernel/sched/core.c file (values from this array should be scaled
1547appropriately so the neutral - nice 0 - value is 100 instead of 1024).
1548
1549
1550IO controller root cgroup process behaviour
1551~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1552
1553Root cgroup processes are hosted in an implicit leaf child node.
1554When distributing IO resources this implicit child node is taken into
1555account as if it was a normal child cgroup of the root cgroup with a
1556weight value of 200.
1557
1558
633b11be
MCC
1559Namespace
1560=========
d4021f6c 1561
633b11be
MCC
1562Basics
1563------
d4021f6c
SH
1564
1565cgroup namespace provides a mechanism to virtualize the view of the
1566"/proc/$PID/cgroup" file and cgroup mounts. The CLONE_NEWCGROUP clone
1567flag can be used with clone(2) and unshare(2) to create a new cgroup
1568namespace. The process running inside the cgroup namespace will have
1569its "/proc/$PID/cgroup" output restricted to cgroupns root. The
1570cgroupns root is the cgroup of the process at the time of creation of
1571the cgroup namespace.
1572
1573Without cgroup namespace, the "/proc/$PID/cgroup" file shows the
1574complete path of the cgroup of a process. In a container setup where
1575a set of cgroups and namespaces are intended to isolate processes the
1576"/proc/$PID/cgroup" file may leak potential system level information
633b11be 1577to the isolated processes. For Example::
d4021f6c
SH
1578
1579 # cat /proc/self/cgroup
1580 0::/batchjobs/container_id1
1581
1582The path '/batchjobs/container_id1' can be considered as system-data
1583and undesirable to expose to the isolated processes. cgroup namespace
1584can be used to restrict visibility of this path. For example, before
633b11be 1585creating a cgroup namespace, one would see::
d4021f6c
SH
1586
1587 # ls -l /proc/self/ns/cgroup
1588 lrwxrwxrwx 1 root root 0 2014-07-15 10:37 /proc/self/ns/cgroup -> cgroup:[4026531835]
1589 # cat /proc/self/cgroup
1590 0::/batchjobs/container_id1
1591
633b11be 1592After unsharing a new namespace, the view changes::
d4021f6c
SH
1593
1594 # ls -l /proc/self/ns/cgroup
1595 lrwxrwxrwx 1 root root 0 2014-07-15 10:35 /proc/self/ns/cgroup -> cgroup:[4026532183]
1596 # cat /proc/self/cgroup
1597 0::/
1598
1599When some thread from a multi-threaded process unshares its cgroup
1600namespace, the new cgroupns gets applied to the entire process (all
1601the threads). This is natural for the v2 hierarchy; however, for the
1602legacy hierarchies, this may be unexpected.
1603
1604A cgroup namespace is alive as long as there are processes inside or
1605mounts pinning it. When the last usage goes away, the cgroup
1606namespace is destroyed. The cgroupns root and the actual cgroups
1607remain.
1608
1609
633b11be
MCC
1610The Root and Views
1611------------------
d4021f6c
SH
1612
1613The 'cgroupns root' for a cgroup namespace is the cgroup in which the
1614process calling unshare(2) is running. For example, if a process in
1615/batchjobs/container_id1 cgroup calls unshare, cgroup
1616/batchjobs/container_id1 becomes the cgroupns root. For the
1617init_cgroup_ns, this is the real root ('/') cgroup.
1618
1619The cgroupns root cgroup does not change even if the namespace creator
633b11be 1620process later moves to a different cgroup::
d4021f6c
SH
1621
1622 # ~/unshare -c # unshare cgroupns in some cgroup
1623 # cat /proc/self/cgroup
1624 0::/
1625 # mkdir sub_cgrp_1
1626 # echo 0 > sub_cgrp_1/cgroup.procs
1627 # cat /proc/self/cgroup
1628 0::/sub_cgrp_1
1629
1630Each process gets its namespace-specific view of "/proc/$PID/cgroup"
1631
1632Processes running inside the cgroup namespace will be able to see
1633cgroup paths (in /proc/self/cgroup) only inside their root cgroup.
633b11be 1634From within an unshared cgroupns::
d4021f6c
SH
1635
1636 # sleep 100000 &
1637 [1] 7353
1638 # echo 7353 > sub_cgrp_1/cgroup.procs
1639 # cat /proc/7353/cgroup
1640 0::/sub_cgrp_1
1641
1642From the initial cgroup namespace, the real cgroup path will be
633b11be 1643visible::
d4021f6c
SH
1644
1645 $ cat /proc/7353/cgroup
1646 0::/batchjobs/container_id1/sub_cgrp_1
1647
1648From a sibling cgroup namespace (that is, a namespace rooted at a
1649different cgroup), the cgroup path relative to its own cgroup
1650namespace root will be shown. For instance, if PID 7353's cgroup
633b11be 1651namespace root is at '/batchjobs/container_id2', then it will see::
d4021f6c
SH
1652
1653 # cat /proc/7353/cgroup
1654 0::/../container_id2/sub_cgrp_1
1655
1656Note that the relative path always starts with '/' to indicate that
1657its relative to the cgroup namespace root of the caller.
1658
1659
633b11be
MCC
1660Migration and setns(2)
1661----------------------
d4021f6c
SH
1662
1663Processes inside a cgroup namespace can move into and out of the
1664namespace root if they have proper access to external cgroups. For
1665example, from inside a namespace with cgroupns root at
1666/batchjobs/container_id1, and assuming that the global hierarchy is
633b11be 1667still accessible inside cgroupns::
d4021f6c
SH
1668
1669 # cat /proc/7353/cgroup
1670 0::/sub_cgrp_1
1671 # echo 7353 > batchjobs/container_id2/cgroup.procs
1672 # cat /proc/7353/cgroup
1673 0::/../container_id2
1674
1675Note that this kind of setup is not encouraged. A task inside cgroup
1676namespace should only be exposed to its own cgroupns hierarchy.
1677
1678setns(2) to another cgroup namespace is allowed when:
1679
1680(a) the process has CAP_SYS_ADMIN against its current user namespace
1681(b) the process has CAP_SYS_ADMIN against the target cgroup
1682 namespace's userns
1683
1684No implicit cgroup changes happen with attaching to another cgroup
1685namespace. It is expected that the someone moves the attaching
1686process under the target cgroup namespace root.
1687
1688
633b11be
MCC
1689Interaction with Other Namespaces
1690---------------------------------
d4021f6c
SH
1691
1692Namespace specific cgroup hierarchy can be mounted by a process
633b11be 1693running inside a non-init cgroup namespace::
d4021f6c
SH
1694
1695 # mount -t cgroup2 none $MOUNT_POINT
1696
1697This will mount the unified cgroup hierarchy with cgroupns root as the
1698filesystem root. The process needs CAP_SYS_ADMIN against its user and
1699mount namespaces.
1700
1701The virtualization of /proc/self/cgroup file combined with restricting
1702the view of cgroup hierarchy by namespace-private cgroupfs mount
1703provides a properly isolated cgroup view inside the container.
1704
1705
633b11be
MCC
1706Information on Kernel Programming
1707=================================
6c292092
TH
1708
1709This section contains kernel programming information in the areas
1710where interacting with cgroup is necessary. cgroup core and
1711controllers are not covered.
1712
1713
633b11be
MCC
1714Filesystem Support for Writeback
1715--------------------------------
6c292092
TH
1716
1717A filesystem can support cgroup writeback by updating
1718address_space_operations->writepage[s]() to annotate bio's using the
1719following two functions.
1720
1721 wbc_init_bio(@wbc, @bio)
6c292092
TH
1722 Should be called for each bio carrying writeback data and
1723 associates the bio with the inode's owner cgroup. Can be
1724 called anytime between bio allocation and submission.
1725
1726 wbc_account_io(@wbc, @page, @bytes)
6c292092
TH
1727 Should be called for each data segment being written out.
1728 While this function doesn't care exactly when it's called
1729 during the writeback session, it's the easiest and most
1730 natural to call it as data segments are added to a bio.
1731
1732With writeback bio's annotated, cgroup support can be enabled per
1733super_block by setting SB_I_CGROUPWB in ->s_iflags. This allows for
1734selective disabling of cgroup writeback support which is helpful when
1735certain filesystem features, e.g. journaled data mode, are
1736incompatible.
1737
1738wbc_init_bio() binds the specified bio to its cgroup. Depending on
1739the configuration, the bio may be executed at a lower priority and if
1740the writeback session is holding shared resources, e.g. a journal
1741entry, may lead to priority inversion. There is no one easy solution
1742for the problem. Filesystems can try to work around specific problem
1743cases by skipping wbc_init_bio() or using bio_associate_blkcg()
1744directly.
1745
1746
633b11be
MCC
1747Deprecated v1 Core Features
1748===========================
6c292092
TH
1749
1750- Multiple hierarchies including named ones are not supported.
1751
5136f636 1752- All v1 mount options are not supported.
6c292092
TH
1753
1754- The "tasks" file is removed and "cgroup.procs" is not sorted.
1755
1756- "cgroup.clone_children" is removed.
1757
1758- /proc/cgroups is meaningless for v2. Use "cgroup.controllers" file
1759 at the root instead.
1760
1761
633b11be
MCC
1762Issues with v1 and Rationales for v2
1763====================================
6c292092 1764
633b11be
MCC
1765Multiple Hierarchies
1766--------------------
6c292092
TH
1767
1768cgroup v1 allowed an arbitrary number of hierarchies and each
1769hierarchy could host any number of controllers. While this seemed to
1770provide a high level of flexibility, it wasn't useful in practice.
1771
1772For example, as there is only one instance of each controller, utility
1773type controllers such as freezer which can be useful in all
1774hierarchies could only be used in one. The issue is exacerbated by
1775the fact that controllers couldn't be moved to another hierarchy once
1776hierarchies were populated. Another issue was that all controllers
1777bound to a hierarchy were forced to have exactly the same view of the
1778hierarchy. It wasn't possible to vary the granularity depending on
1779the specific controller.
1780
1781In practice, these issues heavily limited which controllers could be
1782put on the same hierarchy and most configurations resorted to putting
1783each controller on its own hierarchy. Only closely related ones, such
1784as the cpu and cpuacct controllers, made sense to be put on the same
1785hierarchy. This often meant that userland ended up managing multiple
1786similar hierarchies repeating the same steps on each hierarchy
1787whenever a hierarchy management operation was necessary.
1788
1789Furthermore, support for multiple hierarchies came at a steep cost.
1790It greatly complicated cgroup core implementation but more importantly
1791the support for multiple hierarchies restricted how cgroup could be
1792used in general and what controllers was able to do.
1793
1794There was no limit on how many hierarchies there might be, which meant
1795that a thread's cgroup membership couldn't be described in finite
1796length. The key might contain any number of entries and was unlimited
1797in length, which made it highly awkward to manipulate and led to
1798addition of controllers which existed only to identify membership,
1799which in turn exacerbated the original problem of proliferating number
1800of hierarchies.
1801
1802Also, as a controller couldn't have any expectation regarding the
1803topologies of hierarchies other controllers might be on, each
1804controller had to assume that all other controllers were attached to
1805completely orthogonal hierarchies. This made it impossible, or at
1806least very cumbersome, for controllers to cooperate with each other.
1807
1808In most use cases, putting controllers on hierarchies which are
1809completely orthogonal to each other isn't necessary. What usually is
1810called for is the ability to have differing levels of granularity
1811depending on the specific controller. In other words, hierarchy may
1812be collapsed from leaf towards root when viewed from specific
1813controllers. For example, a given configuration might not care about
1814how memory is distributed beyond a certain level while still wanting
1815to control how CPU cycles are distributed.
1816
1817
633b11be
MCC
1818Thread Granularity
1819------------------
6c292092
TH
1820
1821cgroup v1 allowed threads of a process to belong to different cgroups.
1822This didn't make sense for some controllers and those controllers
1823ended up implementing different ways to ignore such situations but
1824much more importantly it blurred the line between API exposed to
1825individual applications and system management interface.
1826
1827Generally, in-process knowledge is available only to the process
1828itself; thus, unlike service-level organization of processes,
1829categorizing threads of a process requires active participation from
1830the application which owns the target process.
1831
1832cgroup v1 had an ambiguously defined delegation model which got abused
1833in combination with thread granularity. cgroups were delegated to
1834individual applications so that they can create and manage their own
1835sub-hierarchies and control resource distributions along them. This
1836effectively raised cgroup to the status of a syscall-like API exposed
1837to lay programs.
1838
1839First of all, cgroup has a fundamentally inadequate interface to be
1840exposed this way. For a process to access its own knobs, it has to
1841extract the path on the target hierarchy from /proc/self/cgroup,
1842construct the path by appending the name of the knob to the path, open
1843and then read and/or write to it. This is not only extremely clunky
1844and unusual but also inherently racy. There is no conventional way to
1845define transaction across the required steps and nothing can guarantee
1846that the process would actually be operating on its own sub-hierarchy.
1847
1848cgroup controllers implemented a number of knobs which would never be
1849accepted as public APIs because they were just adding control knobs to
1850system-management pseudo filesystem. cgroup ended up with interface
1851knobs which were not properly abstracted or refined and directly
1852revealed kernel internal details. These knobs got exposed to
1853individual applications through the ill-defined delegation mechanism
1854effectively abusing cgroup as a shortcut to implementing public APIs
1855without going through the required scrutiny.
1856
1857This was painful for both userland and kernel. Userland ended up with
1858misbehaving and poorly abstracted interfaces and kernel exposing and
1859locked into constructs inadvertently.
1860
1861
633b11be
MCC
1862Competition Between Inner Nodes and Threads
1863-------------------------------------------
6c292092
TH
1864
1865cgroup v1 allowed threads to be in any cgroups which created an
1866interesting problem where threads belonging to a parent cgroup and its
1867children cgroups competed for resources. This was nasty as two
1868different types of entities competed and there was no obvious way to
1869settle it. Different controllers did different things.
1870
1871The cpu controller considered threads and cgroups as equivalents and
1872mapped nice levels to cgroup weights. This worked for some cases but
1873fell flat when children wanted to be allocated specific ratios of CPU
1874cycles and the number of internal threads fluctuated - the ratios
1875constantly changed as the number of competing entities fluctuated.
1876There also were other issues. The mapping from nice level to weight
1877wasn't obvious or universal, and there were various other knobs which
1878simply weren't available for threads.
1879
1880The io controller implicitly created a hidden leaf node for each
1881cgroup to host the threads. The hidden leaf had its own copies of all
633b11be 1882the knobs with ``leaf_`` prefixed. While this allowed equivalent
6c292092
TH
1883control over internal threads, it was with serious drawbacks. It
1884always added an extra layer of nesting which wouldn't be necessary
1885otherwise, made the interface messy and significantly complicated the
1886implementation.
1887
1888The memory controller didn't have a way to control what happened
1889between internal tasks and child cgroups and the behavior was not
1890clearly defined. There were attempts to add ad-hoc behaviors and
1891knobs to tailor the behavior to specific workloads which would have
1892led to problems extremely difficult to resolve in the long term.
1893
1894Multiple controllers struggled with internal tasks and came up with
1895different ways to deal with it; unfortunately, all the approaches were
1896severely flawed and, furthermore, the widely different behaviors
1897made cgroup as a whole highly inconsistent.
1898
1899This clearly is a problem which needs to be addressed from cgroup core
1900in a uniform way.
1901
1902
633b11be
MCC
1903Other Interface Issues
1904----------------------
6c292092
TH
1905
1906cgroup v1 grew without oversight and developed a large number of
1907idiosyncrasies and inconsistencies. One issue on the cgroup core side
1908was how an empty cgroup was notified - a userland helper binary was
1909forked and executed for each event. The event delivery wasn't
1910recursive or delegatable. The limitations of the mechanism also led
1911to in-kernel event delivery filtering mechanism further complicating
1912the interface.
1913
1914Controller interfaces were problematic too. An extreme example is
1915controllers completely ignoring hierarchical organization and treating
1916all cgroups as if they were all located directly under the root
1917cgroup. Some controllers exposed a large amount of inconsistent
1918implementation details to userland.
1919
1920There also was no consistency across controllers. When a new cgroup
1921was created, some controllers defaulted to not imposing extra
1922restrictions while others disallowed any resource usage until
1923explicitly configured. Configuration knobs for the same type of
1924control used widely differing naming schemes and formats. Statistics
1925and information knobs were named arbitrarily and used different
1926formats and units even in the same controller.
1927
1928cgroup v2 establishes common conventions where appropriate and updates
1929controllers so that they expose minimal and consistent interfaces.
1930
1931
633b11be
MCC
1932Controller Issues and Remedies
1933------------------------------
6c292092 1934
633b11be
MCC
1935Memory
1936~~~~~~
6c292092
TH
1937
1938The original lower boundary, the soft limit, is defined as a limit
1939that is per default unset. As a result, the set of cgroups that
1940global reclaim prefers is opt-in, rather than opt-out. The costs for
1941optimizing these mostly negative lookups are so high that the
1942implementation, despite its enormous size, does not even provide the
1943basic desirable behavior. First off, the soft limit has no
1944hierarchical meaning. All configured groups are organized in a global
1945rbtree and treated like equal peers, regardless where they are located
1946in the hierarchy. This makes subtree delegation impossible. Second,
1947the soft limit reclaim pass is so aggressive that it not just
1948introduces high allocation latencies into the system, but also impacts
1949system performance due to overreclaim, to the point where the feature
1950becomes self-defeating.
1951
1952The memory.low boundary on the other hand is a top-down allocated
1953reserve. A cgroup enjoys reclaim protection when it and all its
1954ancestors are below their low boundaries, which makes delegation of
1955subtrees possible. Secondly, new cgroups have no reserve per default
1956and in the common case most cgroups are eligible for the preferred
1957reclaim pass. This allows the new low boundary to be efficiently
1958implemented with just a minor addition to the generic reclaim code,
1959without the need for out-of-band data structures and reclaim passes.
1960Because the generic reclaim code considers all cgroups except for the
1961ones running low in the preferred first reclaim pass, overreclaim of
1962individual groups is eliminated as well, resulting in much better
1963overall workload performance.
1964
1965The original high boundary, the hard limit, is defined as a strict
1966limit that can not budge, even if the OOM killer has to be called.
1967But this generally goes against the goal of making the most out of the
1968available memory. The memory consumption of workloads varies during
1969runtime, and that requires users to overcommit. But doing that with a
1970strict upper limit requires either a fairly accurate prediction of the
1971working set size or adding slack to the limit. Since working set size
1972estimation is hard and error prone, and getting it wrong results in
1973OOM kills, most users tend to err on the side of a looser limit and
1974end up wasting precious resources.
1975
1976The memory.high boundary on the other hand can be set much more
1977conservatively. When hit, it throttles allocations by forcing them
1978into direct reclaim to work off the excess, but it never invokes the
1979OOM killer. As a result, a high boundary that is chosen too
1980aggressively will not terminate the processes, but instead it will
1981lead to gradual performance degradation. The user can monitor this
1982and make corrections until the minimal memory footprint that still
1983gives acceptable performance is found.
1984
1985In extreme cases, with many concurrent allocations and a complete
1986breakdown of reclaim progress within the group, the high boundary can
1987be exceeded. But even then it's mostly better to satisfy the
1988allocation from the slack available in other groups or the rest of the
1989system than killing the group. Otherwise, memory.max is there to
1990limit this type of spillover and ultimately contain buggy or even
1991malicious applications.
3e24b19d 1992
b6e6edcf
JW
1993Setting the original memory.limit_in_bytes below the current usage was
1994subject to a race condition, where concurrent charges could cause the
1995limit setting to fail. memory.max on the other hand will first set the
1996limit to prevent new charges, and then reclaim and OOM kill until the
1997new limit is met - or the task writing to memory.max is killed.
1998
3e24b19d
VD
1999The combined memory+swap accounting and limiting is replaced by real
2000control over swap space.
2001
2002The main argument for a combined memory+swap facility in the original
2003cgroup design was that global or parental pressure would always be
2004able to swap all anonymous memory of a child group, regardless of the
2005child's own (possibly untrusted) configuration. However, untrusted
2006groups can sabotage swapping by other means - such as referencing its
2007anonymous memory in a tight loop - and an admin can not assume full
2008swappability when overcommitting untrusted jobs.
2009
2010For trusted jobs, on the other hand, a combined counter is not an
2011intuitive userspace interface, and it flies in the face of the idea
2012that cgroup controllers should account and limit specific physical
2013resources. Swap space is a resource like all others in the system,
2014and that's why unified hierarchy allows distributing it separately.