1 ==========================
2 Memory Resource Controller
3 ==========================
4
5 .. caution::
6 This document is hopelessly outdated and it asks for a complete
7 rewrite. It still contains a useful information so we are keeping it
8 here but make sure to check the current code if you need a deeper
9 understanding.
10
11 .. note::
12 The Memory Resource Controller has generically been referred to as the
13 memory controller in this document. Do not confuse memory controller
14 used here with the memory controller that is used in hardware.
15
16 .. hint::
17 When we mention a cgroup (cgroupfs's directory) with memory controller,
18 we call it "memory cgroup". When you see git-log and source code, you'll
19 see patch's title and function names tend to use "memcg".
20 In this document, we avoid using it.
21
22 Benefits and Purpose of the memory controller
23 =============================================
24
25 The memory controller isolates the memory behaviour of a group of tasks
26 from the rest of the system. The article on LWN [12]_ mentions some probable
27 uses of the memory controller. The memory controller can be used to
28
29 a. Isolate an application or a group of applications
30 Memory-hungry applications can be isolated and limited to a smaller
31 amount of memory.
32 b. Create a cgroup with a limited amount of memory; this can be used
33 as a good alternative to booting with mem=XXXX.
34 c. Virtualization solutions can control the amount of memory they want
35 to assign to a virtual machine instance.
36 d. A CD/DVD burner could control the amount of memory used by the
37 rest of the system to ensure that burning does not fail due to lack
38 of available memory.
39 e. There are several other use cases; find one or use the controller just
40 for fun (to learn and hack on the VM subsystem).
41
42 Current Status: linux-2.6.34-mmotm(development version of 2010/April)
43
44 Features:
45
46 - accounting anonymous pages, file caches, swap caches usage and limiting them.
47 - pages are linked to per-memcg LRU exclusively, and there is no global LRU.
48 - optionally, memory+swap usage can be accounted and limited.
49 - hierarchical accounting
50 - soft limit
51 - moving (recharging) account at moving a task is selectable.
52 - usage threshold notifier
53 - memory pressure notifier
54 - oom-killer disable knob and oom-notifier
55 - Root cgroup has no limit controls.
56
57 Kernel memory support is a work in progress, and the current version provides
58 basically functionality. (See :ref:`section 2.7
59 <cgroup-v1-memory-kernel-extension>`)
60
61 Brief summary of control files.
62
63 ==================================== ==========================================
64 tasks attach a task(thread) and show list of
65 threads
66 cgroup.procs show list of processes
67 cgroup.event_control an interface for event_fd()
68 This knob is not available on CONFIG_PREEMPT_RT systems.
69 memory.usage_in_bytes show current usage for memory
70 (See 5.5 for details)
71 memory.memsw.usage_in_bytes show current usage for memory+Swap
72 (See 5.5 for details)
73 memory.limit_in_bytes set/show limit of memory usage
74 memory.memsw.limit_in_bytes set/show limit of memory+Swap usage
75 memory.failcnt show the number of memory usage hits limits
76 memory.memsw.failcnt show the number of memory+Swap hits limits
77 memory.max_usage_in_bytes show max memory usage recorded
78 memory.memsw.max_usage_in_bytes show max memory+Swap usage recorded
79 memory.soft_limit_in_bytes set/show soft limit of memory usage
80 This knob is not available on CONFIG_PREEMPT_RT systems.
81 memory.stat show various statistics
82 memory.use_hierarchy set/show hierarchical account enabled
83 This knob is deprecated and shouldn't be
84 used.
85 memory.force_empty trigger forced page reclaim
86 memory.pressure_level set memory pressure notifications
87 memory.swappiness set/show swappiness parameter of vmscan
88 (See sysctl's vm.swappiness)
89 memory.move_charge_at_immigrate set/show controls of moving charges
90 This knob is deprecated and shouldn't be
91 used.
92 memory.oom_control set/show oom controls.
93 memory.numa_stat show the number of memory usage per numa
94 node
95 memory.kmem.limit_in_bytes Deprecated knob to set and read the kernel
96 memory hard limit. Kernel hard limit is not
97 supported since 5.16. Writing any value to
98 do file will not have any effect same as if
99 nokmem kernel parameter was specified.
100 Kernel memory is still charged and reported
101 by memory.kmem.usage_in_bytes.
102 memory.kmem.usage_in_bytes show current kernel memory allocation
103 memory.kmem.failcnt show the number of kernel memory usage
104 hits limits
105 memory.kmem.max_usage_in_bytes show max kernel memory usage recorded
106
107 memory.kmem.tcp.limit_in_bytes set/show hard limit for tcp buf memory
108 memory.kmem.tcp.usage_in_bytes show current tcp buf memory allocation
109 memory.kmem.tcp.failcnt show the number of tcp buf memory usage
110 hits limits
111 memory.kmem.tcp.max_usage_in_bytes show max tcp buf memory usage recorded
112 ==================================== ==========================================
113
114 1. History
115 ==========
116
117 The memory controller has a long history. A request for comments for the memory
118 controller was posted by Balbir Singh [1]_. At the time the RFC was posted
119 there were several implementations for memory control. The goal of the
120 RFC was to build consensus and agreement for the minimal features required
121 for memory control. The first RSS controller was posted by Balbir Singh [2]_
122 in Feb 2007. Pavel Emelianov [3]_ [4]_ [5]_ has since posted three versions
123 of the RSS controller. At OLS, at the resource management BoF, everyone
124 suggested that we handle both page cache and RSS together. Another request was
125 raised to allow user space handling of OOM. The current memory controller is
126 at version 6; it combines both mapped (RSS) and unmapped Page
127 Cache Control [11]_.
128
129 2. Memory Control
130 =================
131
132 Memory is a unique resource in the sense that it is present in a limited
133 amount. If a task requires a lot of CPU processing, the task can spread
134 its processing over a period of hours, days, months or years, but with
135 memory, the same physical memory needs to be reused to accomplish the task.
136
137 The memory controller implementation has been divided into phases. These
138 are:
139
140 1. Memory controller
141 2. mlock(2) controller
142 3. Kernel user memory accounting and slab control
143 4. user mappings length controller
144
145 The memory controller is the first controller developed.
146
147 2.1. Design
148 -----------
149
150 The core of the design is a counter called the page_counter. The
151 page_counter tracks the current memory usage and limit of the group of
152 processes associated with the controller. Each cgroup has a memory controller
153 specific data structure (mem_cgroup) associated with it.
154
155 2.2. Accounting
156 ---------------
157
158 .. code-block::
159 :caption: Figure 1: Hierarchy of Accounting
160
161 +--------------------+
162 | mem_cgroup |
163 | (page_counter) |
164 +--------------------+
165 / ^ \
166 / | \
167 +---------------+ | +---------------+
168 | mm_struct | |.... | mm_struct |
169 | | | | |
170 +---------------+ | +---------------+
171 |
172 + --------------+
173 |
174 +---------------+ +------+--------+
175 | page +----------> page_cgroup|
176 | | | |
177 +---------------+ +---------------+
178
179
180
181 Figure 1 shows the important aspects of the controller
182
183 1. Accounting happens per cgroup
184 2. Each mm_struct knows about which cgroup it belongs to
185 3. Each page has a pointer to the page_cgroup, which in turn knows the
186 cgroup it belongs to
187
188 The accounting is done as follows: mem_cgroup_charge_common() is invoked to
189 set up the necessary data structures and check if the cgroup that is being
190 charged is over its limit. If it is, then reclaim is invoked on the cgroup.
191 More details can be found in the reclaim section of this document.
192 If everything goes well, a page meta-data-structure called page_cgroup is
193 updated. page_cgroup has its own LRU on cgroup.
194 (*) page_cgroup structure is allocated at boot/memory-hotplug time.
195
196 2.2.1 Accounting details
197 ------------------------
198
199 All mapped anon pages (RSS) and cache pages (Page Cache) are accounted.
200 Some pages which are never reclaimable and will not be on the LRU
201 are not accounted. We just account pages under usual VM management.
202
203 RSS pages are accounted at page_fault unless they've already been accounted
204 for earlier. A file page will be accounted for as Page Cache when it's
205 inserted into inode (xarray). While it's mapped into the page tables of
206 processes, duplicate accounting is carefully avoided.
207
208 An RSS page is unaccounted when it's fully unmapped. A PageCache page is
209 unaccounted when it's removed from xarray. Even if RSS pages are fully
210 unmapped (by kswapd), they may exist as SwapCache in the system until they
211 are really freed. Such SwapCaches are also accounted.
212 A swapped-in page is accounted after adding into swapcache.
213
214 Note: The kernel does swapin-readahead and reads multiple swaps at once.
215 Since page's memcg recorded into swap whatever memsw enabled, the page will
216 be accounted after swapin.
217
218 At page migration, accounting information is kept.
219
220 Note: we just account pages-on-LRU because our purpose is to control amount
221 of used pages; not-on-LRU pages tend to be out-of-control from VM view.
222
223 2.3 Shared Page Accounting
224 --------------------------
225
226 Shared pages are accounted on the basis of the first touch approach. The
227 cgroup that first touches a page is accounted for the page. The principle
228 behind this approach is that a cgroup that aggressively uses a shared
229 page will eventually get charged for it (once it is uncharged from
230 the cgroup that brought it in -- this will happen on memory pressure).
231
232 But see :ref:`section 8.2 <cgroup-v1-memory-movable-charges>` when moving a
233 task to another cgroup, its pages may be recharged to the new cgroup, if
234 move_charge_at_immigrate has been chosen.
235
236 2.4 Swap Extension
237 --------------------------------------
238
239 Swap usage is always recorded for each of cgroup. Swap Extension allows you to
240 read and limit it.
241
242 When CONFIG_SWAP is enabled, following files are added.
243
244 - memory.memsw.usage_in_bytes.
245 - memory.memsw.limit_in_bytes.
246
247 memsw means memory+swap. Usage of memory+swap is limited by
248 memsw.limit_in_bytes.
249
250 Example: Assume a system with 4G of swap. A task which allocates 6G of memory
251 (by mistake) under 2G memory limitation will use all swap.
252 In this case, setting memsw.limit_in_bytes=3G will prevent bad use of swap.
253 By using the memsw limit, you can avoid system OOM which can be caused by swap
254 shortage.
255
256 2.4.1 why 'memory+swap' rather than swap
257 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
258
259 The global LRU(kswapd) can swap out arbitrary pages. Swap-out means
260 to move account from memory to swap...there is no change in usage of
261 memory+swap. In other words, when we want to limit the usage of swap without
262 affecting global LRU, memory+swap limit is better than just limiting swap from
263 an OS point of view.
264
265 2.4.2. What happens when a cgroup hits memory.memsw.limit_in_bytes
266 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
267
268 When a cgroup hits memory.memsw.limit_in_bytes, it's useless to do swap-out
269 in this cgroup. Then, swap-out will not be done by cgroup routine and file
270 caches are dropped. But as mentioned above, global LRU can do swapout memory
271 from it for sanity of the system's memory management state. You can't forbid
272 it by cgroup.
273
274 2.5 Reclaim
275 -----------
276
277 Each cgroup maintains a per cgroup LRU which has the same structure as
278 global VM. When a cgroup goes over its limit, we first try
279 to reclaim memory from the cgroup so as to make space for the new
280 pages that the cgroup has touched. If the reclaim is unsuccessful,
281 an OOM routine is invoked to select and kill the bulkiest task in the
282 cgroup. (See :ref:`10. OOM Control <cgroup-v1-memory-oom-control>` below.)
283
284 The reclaim algorithm has not been modified for cgroups, except that
285 pages that are selected for reclaiming come from the per-cgroup LRU
286 list.
287
288 .. note::
289 Reclaim does not work for the root cgroup, since we cannot set any
290 limits on the root cgroup.
291
292 .. note::
293 When panic_on_oom is set to "2", the whole system will panic.
294
295 When oom event notifier is registered, event will be delivered.
296 (See :ref:`oom_control <cgroup-v1-memory-oom-control>` section)
297
298 2.6 Locking
299 -----------
300
301 Lock order is as follows::
302
303 folio_lock
304 mm->page_table_lock or split pte_lock
305 folio_memcg_lock (memcg->move_lock)
306 mapping->i_pages lock
307 lruvec->lru_lock.
308
309 Per-node-per-memcgroup LRU (cgroup's private LRU) is guarded by
310 lruvec->lru_lock; the folio LRU flag is cleared before
311 isolating a page from its LRU under lruvec->lru_lock.
312
313 .. _cgroup-v1-memory-kernel-extension:
314
315 2.7 Kernel Memory Extension
316 -----------------------------------------------
317
318 With the Kernel memory extension, the Memory Controller is able to limit
319 the amount of kernel memory used by the system. Kernel memory is fundamentally
320 different than user memory, since it can't be swapped out, which makes it
321 possible to DoS the system by consuming too much of this precious resource.
322
323 Kernel memory accounting is enabled for all memory cgroups by default. But
324 it can be disabled system-wide by passing cgroup.memory=nokmem to the kernel
325 at boot time. In this case, kernel memory will not be accounted at all.
326
327 Kernel memory limits are not imposed for the root cgroup. Usage for the root
328 cgroup may or may not be accounted. The memory used is accumulated into
329 memory.kmem.usage_in_bytes, or in a separate counter when it makes sense.
330 (currently only for tcp).
331
332 The main "kmem" counter is fed into the main counter, so kmem charges will
333 also be visible from the user counter.
334
335 Currently no soft limit is implemented for kernel memory. It is future work
336 to trigger slab reclaim when those limits are reached.
337
338 2.7.1 Current Kernel Memory resources accounted
339 -----------------------------------------------
340
341 stack pages:
342 every process consumes some stack pages. By accounting into
343 kernel memory, we prevent new processes from being created when the kernel
344 memory usage is too high.
345
346 slab pages:
347 pages allocated by the SLAB or SLUB allocator are tracked. A copy
348 of each kmem_cache is created every time the cache is touched by the first time
349 from inside the memcg. The creation is done lazily, so some objects can still be
350 skipped while the cache is being created. All objects in a slab page should
351 belong to the same memcg. This only fails to hold when a task is migrated to a
352 different memcg during the page allocation by the cache.
353
354 sockets memory pressure:
355 some sockets protocols have memory pressure
356 thresholds. The Memory Controller allows them to be controlled individually
357 per cgroup, instead of globally.
358
359 tcp memory pressure:
360 sockets memory pressure for the tcp protocol.
361
362 2.7.2 Common use cases
363 ----------------------
364
365 Because the "kmem" counter is fed to the main user counter, kernel memory can
366 never be limited completely independently of user memory. Say "U" is the user
367 limit, and "K" the kernel limit. There are three possible ways limits can be
368 set:
369
370 U != 0, K = unlimited:
371 This is the standard memcg limitation mechanism already present before kmem
372 accounting. Kernel memory is completely ignored.
373
374 U != 0, K < U:
375 Kernel memory is a subset of the user memory. This setup is useful in
376 deployments where the total amount of memory per-cgroup is overcommitted.
377 Overcommitting kernel memory limits is definitely not recommended, since the
378 box can still run out of non-reclaimable memory.
379 In this case, the admin could set up K so that the sum of all groups is
380 never greater than the total memory, and freely set U at the cost of his
381 QoS.
382
383 .. warning::
384 In the current implementation, memory reclaim will NOT be triggered for
385 a cgroup when it hits K while staying below U, which makes this setup
386 impractical.
387
388 U != 0, K >= U:
389 Since kmem charges will also be fed to the user counter and reclaim will be
390 triggered for the cgroup for both kinds of memory. This setup gives the
391 admin a unified view of memory, and it is also useful for people who just
392 want to track kernel memory usage.
393
394 3. User Interface
395 =================
396
397 To use the user interface:
398
399 1. Enable CONFIG_CGROUPS and CONFIG_MEMCG options
400 2. Prepare the cgroups (see :ref:`Why are cgroups needed?
401 <cgroups-why-needed>` for the background information)::
402
403 # mount -t tmpfs none /sys/fs/cgroup
404 # mkdir /sys/fs/cgroup/memory
405 # mount -t cgroup none /sys/fs/cgroup/memory -o memory
406
407 3. Make the new group and move bash into it::
408
409 # mkdir /sys/fs/cgroup/memory/0
410 # echo $$ > /sys/fs/cgroup/memory/0/tasks
411
412 4. Since now we're in the 0 cgroup, we can alter the memory limit::
413
414 # echo 4M > /sys/fs/cgroup/memory/0/memory.limit_in_bytes
415
416 The limit can now be queried::
417
418 # cat /sys/fs/cgroup/memory/0/memory.limit_in_bytes
419 4194304
420
421 .. note::
422 We can use a suffix (k, K, m, M, g or G) to indicate values in kilo,
423 mega or gigabytes. (Here, Kilo, Mega, Giga are Kibibytes, Mebibytes,
424 Gibibytes.)
425
426 .. note::
427 We can write "-1" to reset the ``*.limit_in_bytes(unlimited)``.
428
429 .. note::
430 We cannot set limits on the root cgroup any more.
431
432
433 We can check the usage::
434
435 # cat /sys/fs/cgroup/memory/0/memory.usage_in_bytes
436 1216512
437
438 A successful write to this file does not guarantee a successful setting of
439 this limit to the value written into the file. This can be due to a
440 number of factors, such as rounding up to page boundaries or the total
441 availability of memory on the system. The user is required to re-read
442 this file after a write to guarantee the value committed by the kernel::
443
444 # echo 1 > memory.limit_in_bytes
445 # cat memory.limit_in_bytes
446 4096
447
448 The memory.failcnt field gives the number of times that the cgroup limit was
449 exceeded.
450
451 The memory.stat file gives accounting information. Now, the number of
452 caches, RSS and Active pages/Inactive pages are shown.
453
454 4. Testing
455 ==========
456
457 For testing features and implementation, see memcg_test.txt.
458
459 Performance test is also important. To see pure memory controller's overhead,
460 testing on tmpfs will give you good numbers of small overheads.
461 Example: do kernel make on tmpfs.
462
463 Page-fault scalability is also important. At measuring parallel
464 page fault test, multi-process test may be better than multi-thread
465 test because it has noise of shared objects/status.
466
467 But the above two are testing extreme situations.
468 Trying usual test under memory controller is always helpful.
469
470 .. _cgroup-v1-memory-test-troubleshoot:
471
472 4.1 Troubleshooting
473 -------------------
474
475 Sometimes a user might find that the application under a cgroup is
476 terminated by the OOM killer. There are several causes for this:
477
478 1. The cgroup limit is too low (just too low to do anything useful)
479 2. The user is using anonymous memory and swap is turned off or too low
480
481 A sync followed by echo 1 > /proc/sys/vm/drop_caches will help get rid of
482 some of the pages cached in the cgroup (page cache pages).
483
484 To know what happens, disabling OOM_Kill as per :ref:`"10. OOM Control"
485 <cgroup-v1-memory-oom-control>` (below) and seeing what happens will be
486 helpful.
487
488 .. _cgroup-v1-memory-test-task-migration:
489
490 4.2 Task migration
491 ------------------
492
493 When a task migrates from one cgroup to another, its charge is not
494 carried forward by default. The pages allocated from the original cgroup still
495 remain charged to it, the charge is dropped when the page is freed or
496 reclaimed.
497
498 You can move charges of a task along with task migration.
499 See :ref:`8. "Move charges at task migration" <cgroup-v1-memory-move-charges>`
500
501 4.3 Removing a cgroup
502 ---------------------
503
504 A cgroup can be removed by rmdir, but as discussed in :ref:`sections 4.1
505 <cgroup-v1-memory-test-troubleshoot>` and :ref:`4.2
506 <cgroup-v1-memory-test-task-migration>`, a cgroup might have some charge
507 associated with it, even though all tasks have migrated away from it. (because
508 we charge against pages, not against tasks.)
509
510 We move the stats to parent, and no change on the charge except uncharging
511 from the child.
512
513 Charges recorded in swap information is not updated at removal of cgroup.
514 Recorded information is discarded and a cgroup which uses swap (swapcache)
515 will be charged as a new owner of it.
516
517 5. Misc. interfaces
518 ===================
519
520 5.1 force_empty
521 ---------------
522 memory.force_empty interface is provided to make cgroup's memory usage empty.
523 When writing anything to this::
524
525 # echo 0 > memory.force_empty
526
527 the cgroup will be reclaimed and as many pages reclaimed as possible.
528
529 The typical use case for this interface is before calling rmdir().
530 Though rmdir() offlines memcg, but the memcg may still stay there due to
531 charged file caches. Some out-of-use page caches may keep charged until
532 memory pressure happens. If you want to avoid that, force_empty will be useful.
533
534 5.2 stat file
535 -------------
536
537 memory.stat file includes following statistics:
538
539 * per-memory cgroup local status
540
541 =============== ===============================================================
542 cache # of bytes of page cache memory.
543 rss # of bytes of anonymous and swap cache memory (includes
544 transparent hugepages).
545 rss_huge # of bytes of anonymous transparent hugepages.
546 mapped_file # of bytes of mapped file (includes tmpfs/shmem)
547 pgpgin # of charging events to the memory cgroup. The charging
548 event happens each time a page is accounted as either mapped
549 anon page(RSS) or cache page(Page Cache) to the cgroup.
550 pgpgout # of uncharging events to the memory cgroup. The uncharging
551 event happens each time a page is unaccounted from the
552 cgroup.
553 swap # of bytes of swap usage
554 swapcached # of bytes of swap cached in memory
555 dirty # of bytes that are waiting to get written back to the disk.
556 writeback # of bytes of file/anon cache that are queued for syncing to
557 disk.
558 inactive_anon # of bytes of anonymous and swap cache memory on inactive
559 LRU list.
560 active_anon # of bytes of anonymous and swap cache memory on active
561 LRU list.
562 inactive_file # of bytes of file-backed memory and MADV_FREE anonymous
563 memory (LazyFree pages) on inactive LRU list.
564 active_file # of bytes of file-backed memory on active LRU list.
565 unevictable # of bytes of memory that cannot be reclaimed (mlocked etc).
566 =============== ===============================================================
567
568 * status considering hierarchy (see memory.use_hierarchy settings):
569
570 ========================= ===================================================
571 hierarchical_memory_limit # of bytes of memory limit with regard to
572 hierarchy
573 under which the memory cgroup is
574 hierarchical_memsw_limit # of bytes of memory+swap limit with regard to
575 hierarchy under which memory cgroup is.
576
577 total_<counter> # hierarchical version of <counter>, which in
578 addition to the cgroup's own value includes the
579 sum of all hierarchical children's values of
580 <counter>, i.e. total_cache
581 ========================= ===================================================
582
583 * additional vm parameters (depends on CONFIG_DEBUG_VM):
584
585 ========================= ========================================
586 recent_rotated_anon VM internal parameter. (see mm/vmscan.c)
587 recent_rotated_file VM internal parameter. (see mm/vmscan.c)
588 recent_scanned_anon VM internal parameter. (see mm/vmscan.c)
589 recent_scanned_file VM internal parameter. (see mm/vmscan.c)
590 ========================= ========================================
591
592 .. hint::
593 recent_rotated means recent frequency of LRU rotation.
594 recent_scanned means recent # of scans to LRU.
595 showing for better debug please see the code for meanings.
596
597 .. note::
598 Only anonymous and swap cache memory is listed as part of 'rss' stat.
599 This should not be confused with the true 'resident set size' or the
600 amount of physical memory used by the cgroup.
601
602 'rss + mapped_file" will give you resident set size of cgroup.
603
604 (Note: file and shmem may be shared among other cgroups. In that case,
605 mapped_file is accounted only when the memory cgroup is owner of page
606 cache.)
607
608 5.3 swappiness
609 --------------
610
611 Overrides /proc/sys/vm/swappiness for the particular group. The tunable
612 in the root cgroup corresponds to the global swappiness setting.
613
614 Please note that unlike during the global reclaim, limit reclaim
615 enforces that 0 swappiness really prevents from any swapping even if
616 there is a swap storage available. This might lead to memcg OOM killer
617 if there are no file pages to reclaim.
618
619 5.4 failcnt
620 -----------
621
622 A memory cgroup provides memory.failcnt and memory.memsw.failcnt files.
623 This failcnt(== failure count) shows the number of times that a usage counter
624 hit its limit. When a memory cgroup hits a limit, failcnt increases and
625 memory under it will be reclaimed.
626
627 You can reset failcnt by writing 0 to failcnt file::
628
629 # echo 0 > .../memory.failcnt
630
631 5.5 usage_in_bytes
632 ------------------
633
634 For efficiency, as other kernel components, memory cgroup uses some optimization
635 to avoid unnecessary cacheline false sharing. usage_in_bytes is affected by the
636 method and doesn't show 'exact' value of memory (and swap) usage, it's a fuzz
637 value for efficient access. (Of course, when necessary, it's synchronized.)
638 If you want to know more exact memory usage, you should use RSS+CACHE(+SWAP)
639 value in memory.stat(see 5.2).
640
641 5.6 numa_stat
642 -------------
643
644 This is similar to numa_maps but operates on a per-memcg basis. This is
645 useful for providing visibility into the numa locality information within
646 an memcg since the pages are allowed to be allocated from any physical
647 node. One of the use cases is evaluating application performance by
648 combining this information with the application's CPU allocation.
649
650 Each memcg's numa_stat file includes "total", "file", "anon" and "unevictable"
651 per-node page counts including "hierarchical_<counter>" which sums up all
652 hierarchical children's values in addition to the memcg's own value.
653
654 The output format of memory.numa_stat is::
655
656 total=<total pages> N0=<node 0 pages> N1=<node 1 pages> ...
657 file=<total file pages> N0=<node 0 pages> N1=<node 1 pages> ...
658 anon=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ...
659 unevictable=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ...
660 hierarchical_<counter>=<counter pages> N0=<node 0 pages> N1=<node 1 pages> ...
661
662 The "total" count is sum of file + anon + unevictable.
663
664 6. Hierarchy support
665 ====================
666
667 The memory controller supports a deep hierarchy and hierarchical accounting.
668 The hierarchy is created by creating the appropriate cgroups in the
669 cgroup filesystem. Consider for example, the following cgroup filesystem
670 hierarchy::
671
672 root
673 / | \
674 / | \
675 a b c
676 | \
677 | \
678 d e
679
680 In the diagram above, with hierarchical accounting enabled, all memory
681 usage of e, is accounted to its ancestors up until the root (i.e, c and root).
682 If one of the ancestors goes over its limit, the reclaim algorithm reclaims
683 from the tasks in the ancestor and the children of the ancestor.
684
685 6.1 Hierarchical accounting and reclaim
686 ---------------------------------------
687
688 Hierarchical accounting is enabled by default. Disabling the hierarchical
689 accounting is deprecated. An attempt to do it will result in a failure
690 and a warning printed to dmesg.
691
692 For compatibility reasons writing 1 to memory.use_hierarchy will always pass::
693
694 # echo 1 > memory.use_hierarchy
695
696 7. Soft limits
697 ==============
698
699 Soft limits allow for greater sharing of memory. The idea behind soft limits
700 is to allow control groups to use as much of the memory as needed, provided
701
702 a. There is no memory contention
703 b. They do not exceed their hard limit
704
705 When the system detects memory contention or low memory, control groups
706 are pushed back to their soft limits. If the soft limit of each control
707 group is very high, they are pushed back as much as possible to make
708 sure that one control group does not starve the others of memory.
709
710 Please note that soft limits is a best-effort feature; it comes with
711 no guarantees, but it does its best to make sure that when memory is
712 heavily contended for, memory is allocated based on the soft limit
713 hints/setup. Currently soft limit based reclaim is set up such that
714 it gets invoked from balance_pgdat (kswapd).
715
716 7.1 Interface
717 -------------
718
719 Soft limits can be setup by using the following commands (in this example we
720 assume a soft limit of 256 MiB)::
721
722 # echo 256M > memory.soft_limit_in_bytes
723
724 If we want to change this to 1G, we can at any time use::
725
726 # echo 1G > memory.soft_limit_in_bytes
727
728 .. note::
729 Soft limits take effect over a long period of time, since they involve
730 reclaiming memory for balancing between memory cgroups
731
732 .. note::
733 It is recommended to set the soft limit always below the hard limit,
734 otherwise the hard limit will take precedence.
735
736 .. _cgroup-v1-memory-move-charges:
737
738 8. Move charges at task migration (DEPRECATED!)
739 ===============================================
740
741 THIS IS DEPRECATED!
742
743 It's expensive and unreliable! It's better practice to launch workload
744 tasks directly from inside their target cgroup. Use dedicated workload
745 cgroups to allow fine-grained policy adjustments without having to
746 move physical pages between control domains.
747
748 Users can move charges associated with a task along with task migration, that
749 is, uncharge task's pages from the old cgroup and charge them to the new cgroup.
750 This feature is not supported in !CONFIG_MMU environments because of lack of
751 page tables.
752
753 8.1 Interface
754 -------------
755
756 This feature is disabled by default. It can be enabled (and disabled again) by
757 writing to memory.move_charge_at_immigrate of the destination cgroup.
758
759 If you want to enable it::
760
761 # echo (some positive value) > memory.move_charge_at_immigrate
762
763 .. note::
764 Each bits of move_charge_at_immigrate has its own meaning about what type
765 of charges should be moved. See :ref:`section 8.2
766 <cgroup-v1-memory-movable-charges>` for details.
767
768 .. note::
769 Charges are moved only when you move mm->owner, in other words,
770 a leader of a thread group.
771
772 .. note::
773 If we cannot find enough space for the task in the destination cgroup, we
774 try to make space by reclaiming memory. Task migration may fail if we
775 cannot make enough space.
776
777 .. note::
778 It can take several seconds if you move charges much.
779
780 And if you want disable it again::
781
782 # echo 0 > memory.move_charge_at_immigrate
783
784 .. _cgroup-v1-memory-movable-charges:
785
786 8.2 Type of charges which can be moved
787 --------------------------------------
788
789 Each bit in move_charge_at_immigrate has its own meaning about what type of
790 charges should be moved. But in any case, it must be noted that an account of
791 a page or a swap can be moved only when it is charged to the task's current
792 (old) memory cgroup.
793
794 +---+--------------------------------------------------------------------------+
795 |bit| what type of charges would be moved ? |
796 +===+==========================================================================+
797 | 0 | A charge of an anonymous page (or swap of it) used by the target task. |
798 | | You must enable Swap Extension (see 2.4) to enable move of swap charges. |
799 +---+--------------------------------------------------------------------------+
800 | 1 | A charge of file pages (normal file, tmpfs file (e.g. ipc shared memory) |
801 | | and swaps of tmpfs file) mmapped by the target task. Unlike the case of |
802 | | anonymous pages, file pages (and swaps) in the range mmapped by the task |
803 | | will be moved even if the task hasn't done page fault, i.e. they might |
804 | | not be the task's "RSS", but other task's "RSS" that maps the same file. |
805 | | The mapcount of the page is ignored (the page can be moved independent |
806 | | of the mapcount). You must enable Swap Extension (see 2.4) to |
807 | | enable move of swap charges. |
808 +---+--------------------------------------------------------------------------+
809
810 8.3 TODO
811 --------
812
813 - All of moving charge operations are done under cgroup_mutex. It's not good
814 behavior to hold the mutex too long, so we may need some trick.
815
816 9. Memory thresholds
817 ====================
818
819 Memory cgroup implements memory thresholds using the cgroups notification
820 API (see cgroups.txt). It allows to register multiple memory and memsw
821 thresholds and gets notifications when it crosses.
822
823 To register a threshold, an application must:
824
825 - create an eventfd using eventfd(2);
826 - open memory.usage_in_bytes or memory.memsw.usage_in_bytes;
827 - write string like "<event_fd> <fd of memory.usage_in_bytes> <threshold>" to
828 cgroup.event_control.
829
830 Application will be notified through eventfd when memory usage crosses
831 threshold in any direction.
832
833 It's applicable for root and non-root cgroup.
834
835 .. _cgroup-v1-memory-oom-control:
836
837 10. OOM Control
838 ===============
839
840 memory.oom_control file is for OOM notification and other controls.
841
842 Memory cgroup implements OOM notifier using the cgroup notification
843 API (See cgroups.txt). It allows to register multiple OOM notification
844 delivery and gets notification when OOM happens.
845
846 To register a notifier, an application must:
847
848 - create an eventfd using eventfd(2)
849 - open memory.oom_control file
850 - write string like "<event_fd> <fd of memory.oom_control>" to
851 cgroup.event_control
852
853 The application will be notified through eventfd when OOM happens.
854 OOM notification doesn't work for the root cgroup.
855
856 You can disable the OOM-killer by writing "1" to memory.oom_control file, as:
857
858 #echo 1 > memory.oom_control
859
860 If OOM-killer is disabled, tasks under cgroup will hang/sleep
861 in memory cgroup's OOM-waitqueue when they request accountable memory.
862
863 For running them, you have to relax the memory cgroup's OOM status by
864
865 * enlarge limit or reduce usage.
866
867 To reduce usage,
868
869 * kill some tasks.
870 * move some tasks to other group with account migration.
871 * remove some files (on tmpfs?)
872
873 Then, stopped tasks will work again.
874
875 At reading, current status of OOM is shown.
876
877 - oom_kill_disable 0 or 1
878 (if 1, oom-killer is disabled)
879 - under_oom 0 or 1
880 (if 1, the memory cgroup is under OOM, tasks may be stopped.)
881 - oom_kill integer counter
882 The number of processes belonging to this cgroup killed by any
883 kind of OOM killer.
884
885 11. Memory Pressure
886 ===================
887
888 The pressure level notifications can be used to monitor the memory
889 allocation cost; based on the pressure, applications can implement
890 different strategies of managing their memory resources. The pressure
891 levels are defined as following:
892
893 The "low" level means that the system is reclaiming memory for new
894 allocations. Monitoring this reclaiming activity might be useful for
895 maintaining cache level. Upon notification, the program (typically
896 "Activity Manager") might analyze vmstat and act in advance (i.e.
897 prematurely shutdown unimportant services).
898
899 The "medium" level means that the system is experiencing medium memory
900 pressure, the system might be making swap, paging out active file caches,
901 etc. Upon this event applications may decide to further analyze
902 vmstat/zoneinfo/memcg or internal memory usage statistics and free any
903 resources that can be easily reconstructed or re-read from a disk.
904
905 The "critical" level means that the system is actively thrashing, it is
906 about to out of memory (OOM) or even the in-kernel OOM killer is on its
907 way to trigger. Applications should do whatever they can to help the
908 system. It might be too late to consult with vmstat or any other
909 statistics, so it's advisable to take an immediate action.
910
911 By default, events are propagated upward until the event is handled, i.e. the
912 events are not pass-through. For example, you have three cgroups: A->B->C. Now
913 you set up an event listener on cgroups A, B and C, and suppose group C
914 experiences some pressure. In this situation, only group C will receive the
915 notification, i.e. groups A and B will not receive it. This is done to avoid
916 excessive "broadcasting" of messages, which disturbs the system and which is
917 especially bad if we are low on memory or thrashing. Group B, will receive
918 notification only if there are no event listeners for group C.
919
920 There are three optional modes that specify different propagation behavior:
921
922 - "default": this is the default behavior specified above. This mode is the
923 same as omitting the optional mode parameter, preserved by backwards
924 compatibility.
925
926 - "hierarchy": events always propagate up to the root, similar to the default
927 behavior, except that propagation continues regardless of whether there are
928 event listeners at each level, with the "hierarchy" mode. In the above
929 example, groups A, B, and C will receive notification of memory pressure.
930
931 - "local": events are pass-through, i.e. they only receive notifications when
932 memory pressure is experienced in the memcg for which the notification is
933 registered. In the above example, group C will receive notification if
934 registered for "local" notification and the group experiences memory
935 pressure. However, group B will never receive notification, regardless if
936 there is an event listener for group C or not, if group B is registered for
937 local notification.
938
939 The level and event notification mode ("hierarchy" or "local", if necessary) are
940 specified by a comma-delimited string, i.e. "low,hierarchy" specifies
941 hierarchical, pass-through, notification for all ancestor memcgs. Notification
942 that is the default, non pass-through behavior, does not specify a mode.
943 "medium,local" specifies pass-through notification for the medium level.
944
945 The file memory.pressure_level is only used to setup an eventfd. To
946 register a notification, an application must:
947
948 - create an eventfd using eventfd(2);
949 - open memory.pressure_level;
950 - write string as "<event_fd> <fd of memory.pressure_level> <level[,mode]>"
951 to cgroup.event_control.
952
953 Application will be notified through eventfd when memory pressure is at
954 the specific level (or higher). Read/write operations to
955 memory.pressure_level are no implemented.
956
957 Test:
958
959 Here is a small script example that makes a new cgroup, sets up a
960 memory limit, sets up a notification in the cgroup and then makes child
961 cgroup experience a critical pressure::
962
963 # cd /sys/fs/cgroup/memory/
964 # mkdir foo
965 # cd foo
966 # cgroup_event_listener memory.pressure_level low,hierarchy &
967 # echo 8000000 > memory.limit_in_bytes
968 # echo 8000000 > memory.memsw.limit_in_bytes
969 # echo $$ > tasks
970 # dd if=/dev/zero | read x
971
972 (Expect a bunch of notifications, and eventually, the oom-killer will
973 trigger.)
974
975 12. TODO
976 ========
977
978 1. Make per-cgroup scanner reclaim not-shared pages first
979 2. Teach controller to account for shared-pages
980 3. Start reclamation in the background when the limit is
981 not yet hit but the usage is getting closer
982
983 Summary
984 =======
985
986 Overall, the memory controller has been a stable controller and has been
987 commented and discussed quite extensively in the community.
988
989 References
990 ==========
991
992 .. [1] Singh, Balbir. RFC: Memory Controller, http://lwn.net/Articles/206697/
993 .. [2] Singh, Balbir. Memory Controller (RSS Control),
994 http://lwn.net/Articles/222762/
995 .. [3] Emelianov, Pavel. Resource controllers based on process cgroups
996 https://lore.kernel.org/r/45ED7DEC.7010403@sw.ru
997 .. [4] Emelianov, Pavel. RSS controller based on process cgroups (v2)
998 https://lore.kernel.org/r/461A3010.90403@sw.ru
999 .. [5] Emelianov, Pavel. RSS controller based on process cgroups (v3)
1000 https://lore.kernel.org/r/465D9739.8070209@openvz.org
1001
1002 6. Menage, Paul. Control Groups v10, http://lwn.net/Articles/236032/
1003 7. Vaidyanathan, Srinivasan, Control Groups: Pagecache accounting and control
1004 subsystem (v3), http://lwn.net/Articles/235534/
1005 8. Singh, Balbir. RSS controller v2 test results (lmbench),
1006 https://lore.kernel.org/r/464C95D4.7070806@linux.vnet.ibm.com
1007 9. Singh, Balbir. RSS controller v2 AIM9 results
1008 https://lore.kernel.org/r/464D267A.50107@linux.vnet.ibm.com
1009 10. Singh, Balbir. Memory controller v6 test results,
1010 https://lore.kernel.org/r/20070819094658.654.84837.sendpatchset@balbir-laptop
1011
1012 .. [11] Singh, Balbir. Memory controller introduction (v6),
1013 https://lore.kernel.org/r/20070817084228.26003.12568.sendpatchset@balbir-laptop
1014 .. [12] Corbet, Jonathan, Controlling memory use in cgroups,
1015 http://lwn.net/Articles/243795/
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.