Lines Matching +full:memory +full:- +full:to +full:- +full:memory
2 Memory Resource Controller
8 here but make sure to check the current code if you need a deeper
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.
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".
22 Benefits and Purpose of the memory controller
25 The memory controller isolates the memory behaviour of a group of tasks
27 uses of the memory controller. The memory controller can be used to
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.
40 for fun (to learn and hack on the VM subsystem).
42 Current Status: linux-2.6.34-mmotm(development version of 2010/April)
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.
57 Kernel memory support is a work in progress, and the current version provides
59 <cgroup-v1-memory-kernel-extension>`)
69 memory.usage_in_bytes show current usage for memory
71 memory.memsw.usage_in_bytes show current usage for memory+Swap
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
83 memory.stat show various statistics
84 memory.use_hierarchy set/show hierarchical account enabled
87 memory.force_empty trigger forced page reclaim
88 memory.pressure_level set memory pressure notifications
91 memory.swappiness set/show swappiness parameter of vmscan
93 memory.move_charge_at_immigrate This knob is deprecated.
94 memory.oom_control set/show oom controls.
97 memory.numa_stat show the number of memory usage per numa
99 memory.kmem.limit_in_bytes Deprecated knob to set and read the kernel
100 memory hard limit. Kernel hard limit is not
101 supported since 5.16. Writing any value to
104 Kernel memory is still charged and reported
105 by memory.kmem.usage_in_bytes.
106 memory.kmem.usage_in_bytes show current kernel memory allocation
107 memory.kmem.failcnt show the number of kernel memory usage
109 memory.kmem.max_usage_in_bytes show max kernel memory usage recorded
111 memory.kmem.tcp.limit_in_bytes set/show hard limit for tcp buf memory
114 memory.kmem.tcp.usage_in_bytes show current tcp buf memory allocation
117 memory.kmem.tcp.failcnt show the number of tcp buf memory usage
121 memory.kmem.tcp.max_usage_in_bytes show max tcp buf memory usage recorded
129 The memory controller has a long history. A request for comments for the memory
131 there were several implementations for memory control. The goal of the
132 RFC was to build consensus and agreement for the minimal features required
133 for memory control. The first RSS controller was posted by Balbir Singh [2]_
137 raised to allow user space handling of OOM. The current memory controller is
141 2. Memory Control
144 Memory is a unique resource in the sense that it is present in a limited
147 memory, the same physical memory needs to be reused to accomplish the task.
149 The memory controller implementation has been divided into phases. These
152 1. Memory controller
154 3. Kernel user memory accounting and slab control
157 The memory controller is the first controller developed.
160 -----------
163 page_counter tracks the current memory usage and limit of the group of
164 processes associated with the controller. Each cgroup has a memory controller
168 ---------------
170 .. code-block::
173 +--------------------+
176 +--------------------+
179 +---------------+ | +---------------+
182 +---------------+ | +---------------+
184 + --------------+
186 +---------------+ +------+--------+
187 | page +----------> page_cgroup|
189 +---------------+ +---------------+
196 2. Each mm_struct knows about which cgroup it belongs to
197 3. Each page has a pointer to the page_cgroup, which in turn knows the
198 cgroup it belongs to
200 The accounting is done as follows: mem_cgroup_charge_common() is invoked to
204 If everything goes well, a page meta-data-structure called page_cgroup is
206 (*) page_cgroup structure is allocated at boot/memory-hotplug time.
209 ------------------------
224 A swapped-in page is accounted after adding into swapcache.
226 Note: The kernel does swapin-readahead and reads multiple swaps at once.
232 Note: we just account pages-on-LRU because our purpose is to control amount
233 of used pages; not-on-LRU pages tend to be out-of-control from VM view.
236 --------------------------
242 the cgroup that brought it in -- this will happen on memory pressure).
245 --------------------------------------
247 Swap usage is always recorded for each of cgroup. Swap Extension allows you to
252 - memory.memsw.usage_in_bytes.
253 - memory.memsw.limit_in_bytes.
255 memsw means memory+swap. Usage of memory+swap is limited by
258 Example: Assume a system with 4G of swap. A task which allocates 6G of memory
259 (by mistake) under 2G memory limitation will use all swap.
264 2.4.1 why 'memory+swap' rather than swap
267 The global LRU(kswapd) can swap out arbitrary pages. Swap-out means
268 to move account from memory to swap...there is no change in usage of
269 memory+swap. In other words, when we want to limit the usage of swap without
270 affecting global LRU, memory+swap limit is better than just limiting swap from
273 2.4.2. What happens when a cgroup hits memory.memsw.limit_in_bytes
276 When a cgroup hits memory.memsw.limit_in_bytes, it's useless to do swap-out
277 in this cgroup. Then, swap-out will not be done by cgroup routine and file
278 caches are dropped. But as mentioned above, global LRU can do swapout memory
279 from it for sanity of the system's memory management state. You can't forbid
283 -----------
287 to reclaim memory from the cgroup so as to make space for the new
289 an OOM routine is invoked to select and kill the bulkiest task in the
290 cgroup. (See :ref:`10. OOM Control <cgroup-v1-memory-oom-control>` below.)
293 pages that are selected for reclaiming come from the per-cgroup LRU
301 When panic_on_oom is set to "2", the whole system will panic.
304 (See :ref:`oom_control <cgroup-v1-memory-oom-control>` section)
307 -----------
312 mm->page_table_lock or split pte_lock
313 folio_memcg_lock (memcg->move_lock)
314 mapping->i_pages lock
315 lruvec->lru_lock.
317 Per-node-per-memcgroup LRU (cgroup's private LRU) is guarded by
318 lruvec->lru_lock; the folio LRU flag is cleared before
319 isolating a page from its LRU under lruvec->lru_lock.
321 .. _cgroup-v1-memory-kernel-extension:
323 2.7 Kernel Memory Extension
324 -----------------------------------------------
326 With the Kernel memory extension, the Memory Controller is able to limit
327 the amount of kernel memory used by the system. Kernel memory is fundamentally
328 different than user memory, since it can't be swapped out, which makes it
329 possible to DoS the system by consuming too much of this precious resource.
331 Kernel memory accounting is enabled for all memory cgroups by default. But
332 it can be disabled system-wide by passing cgroup.memory=nokmem to the kernel
333 at boot time. In this case, kernel memory will not be accounted at all.
335 Kernel memory limits are not imposed for the root cgroup. Usage for the root
336 cgroup may or may not be accounted. The memory used is accumulated into
337 memory.kmem.usage_in_bytes, or in a separate counter when it makes sense.
343 Currently no soft limit is implemented for kernel memory. It is future work
344 to trigger slab reclaim when those limits are reached.
346 2.7.1 Current Kernel Memory resources accounted
347 -----------------------------------------------
351 kernel memory, we prevent new processes from being created when the kernel
352 memory usage is too high.
359 belong to the same memcg. This only fails to hold when a task is migrated to a
362 sockets memory pressure:
363 some sockets protocols have memory pressure
364 thresholds. The Memory Controller allows them to be controlled individually
367 tcp memory pressure:
368 sockets memory pressure for the tcp protocol.
371 ----------------------
373 Because the "kmem" counter is fed to the main user counter, kernel memory can
374 never be limited completely independently of user memory. Say "U" is the user
380 accounting. Kernel memory is completely ignored.
383 Kernel memory is a subset of the user memory. This setup is useful in
384 deployments where the total amount of memory per-cgroup is overcommitted.
385 Overcommitting kernel memory limits is definitely not recommended, since the
386 box can still run out of non-reclaimable memory.
388 never greater than the total memory, and freely set U at the cost of his
392 In the current implementation, memory reclaim will NOT be triggered for
397 Since kmem charges will also be fed to the user counter and reclaim will be
398 triggered for the cgroup for both kinds of memory. This setup gives the
399 admin a unified view of memory, and it is also useful for people who just
400 want to track kernel memory usage.
405 To use the user interface:
409 <cgroups-why-needed>` for the background information)::
411 # mount -t tmpfs none /sys/fs/cgroup
412 # mkdir /sys/fs/cgroup/memory
413 # mount -t cgroup none /sys/fs/cgroup/memory -o memory
417 # mkdir /sys/fs/cgroup/memory/0
418 # echo $$ > /sys/fs/cgroup/memory/0/tasks
420 4. Since now we're in the 0 cgroup, we can alter the memory limit::
422 # echo 4M > /sys/fs/cgroup/memory/0/memory.limit_in_bytes
426 # cat /sys/fs/cgroup/memory/0/memory.limit_in_bytes
430 We can use a suffix (k, K, m, M, g or G) to indicate values in kilo,
435 We can write "-1" to reset the ``*.limit_in_bytes(unlimited)``.
443 # cat /sys/fs/cgroup/memory/0/memory.usage_in_bytes
446 A successful write to this file does not guarantee a successful setting of
447 this limit to the value written into the file. This can be due to a
448 number of factors, such as rounding up to page boundaries or the total
449 availability of memory on the system. The user is required to re-read
450 this file after a write to guarantee the value committed by the kernel::
452 # echo 1 > memory.limit_in_bytes
453 # cat memory.limit_in_bytes
456 The memory.failcnt field gives the number of times that the cgroup limit was
459 The memory.stat file gives accounting information. Now, the number of
467 Performance test is also important. To see pure memory controller's overhead,
471 Page-fault scalability is also important. At measuring parallel
472 page fault test, multi-process test may be better than multi-thread
476 Trying usual test under memory controller is always helpful.
478 .. _cgroup-v1-memory-test-troubleshoot:
481 -------------------
486 1. The cgroup limit is too low (just too low to do anything useful)
487 2. The user is using anonymous memory and swap is turned off or too low
492 To know what happens, disabling OOM_Kill as per :ref:`"10. OOM Control"
493 <cgroup-v1-memory-oom-control>` (below) and seeing what happens will be
496 .. _cgroup-v1-memory-test-task-migration:
499 ------------------
501 When a task migrates from one cgroup to another, its charge is not
503 remain charged to it, the charge is dropped when the page is freed or
507 See :ref:`8. "Move charges at task migration" <cgroup-v1-memory-move-charges>`
510 ---------------------
513 <cgroup-v1-memory-test-troubleshoot>` and :ref:`4.2
514 <cgroup-v1-memory-test-task-migration>`, a cgroup might have some charge
518 We move the stats to parent, and no change on the charge except uncharging
529 ---------------
530 memory.force_empty interface is provided to make cgroup's memory usage empty.
531 When writing anything to this::
533 # echo 0 > memory.force_empty
538 Though rmdir() offlines memcg, but the memcg may still stay there due to
539 charged file caches. Some out-of-use page caches may keep charged until
540 memory pressure happens. If you want to avoid that, force_empty will be useful.
543 -------------
545 memory.stat file includes following statistics:
547 * per-memory cgroup local status
550 cache # of bytes of page cache memory.
551 rss # of bytes of anonymous and swap cache memory (includes
555 pgpgin # of charging events to the memory cgroup. The charging
557 anon page(RSS) or cache page(Page Cache) to the cgroup.
558 pgpgout # of uncharging events to the memory cgroup. The uncharging
562 swapcached # of bytes of swap cached in memory
563 dirty # of bytes that are waiting to get written back to the disk.
564 writeback # of bytes of file/anon cache that are queued for syncing to
566 inactive_anon # of bytes of anonymous and swap cache memory on inactive
568 active_anon # of bytes of anonymous and swap cache memory on active
570 inactive_file # of bytes of file-backed memory and MADV_FREE anonymous
571 memory (LazyFree pages) on inactive LRU list.
572 active_file # of bytes of file-backed memory on active LRU list.
573 unevictable # of bytes of memory that cannot be reclaimed (mlocked etc).
576 * status considering hierarchy (see memory.use_hierarchy settings):
579 hierarchical_memory_limit # of bytes of memory limit with regard to
581 under which the memory cgroup is
582 hierarchical_memsw_limit # of bytes of memory+swap limit with regard to
583 hierarchy under which memory cgroup is.
586 addition to the cgroup's own value includes the
602 recent_scanned means recent # of scans to LRU.
606 Only anonymous and swap cache memory is listed as part of 'rss' stat.
608 amount of physical memory used by the cgroup.
613 mapped_file is accounted only when the memory cgroup is owner of page
617 --------------
620 in the root cgroup corresponds to the global swappiness setting.
624 there is a swap storage available. This might lead to memcg OOM killer
625 if there are no file pages to reclaim.
628 -----------
630 A memory cgroup provides memory.failcnt and memory.memsw.failcnt files.
632 hit its limit. When a memory cgroup hits a limit, failcnt increases and
633 memory under it will be reclaimed.
635 You can reset failcnt by writing 0 to failcnt file::
637 # echo 0 > .../memory.failcnt
640 ------------------
642 For efficiency, as other kernel components, memory cgroup uses some optimization
643 to avoid unnecessary cacheline false sharing. usage_in_bytes is affected by the
644 method and doesn't show 'exact' value of memory (and swap) usage, it's a fuzz
646 If you want to know more exact memory usage, you should use RSS+CACHE(+SWAP)
647 value in memory.stat(see 5.2).
650 -------------
652 This is similar to numa_maps but operates on a per-memcg basis. This is
654 an memcg since the pages are allowed to be allocated from any physical
659 per-node page counts including "hierarchical_<counter>" which sums up all
660 hierarchical children's values in addition to the memcg's own value.
662 The output format of memory.numa_stat is::
675 The memory controller supports a deep hierarchy and hierarchical accounting.
688 In the diagram above, with hierarchical accounting enabled, all memory
689 usage of e, is accounted to its ancestors up until the root (i.e, c and root).
694 ---------------------------------------
697 accounting is deprecated. An attempt to do it will result in a failure
698 and a warning printed to dmesg.
700 For compatibility reasons writing 1 to memory.use_hierarchy will always pass::
702 # echo 1 > memory.use_hierarchy
709 Soft limits allow for greater sharing of memory. The idea behind soft limits
710 is to allow control groups to use as much of the memory as needed, provided
712 a. There is no memory contention
715 When the system detects memory contention or low memory, control groups
716 are pushed back to their soft limits. If the soft limit of each control
717 group is very high, they are pushed back as much as possible to make
718 sure that one control group does not starve the others of memory.
720 Please note that soft limits is a best-effort feature; it comes with
721 no guarantees, but it does its best to make sure that when memory is
722 heavily contended for, memory is allocated based on the soft limit
727 -------------
732 # echo 256M > memory.soft_limit_in_bytes
734 If we want to change this to 1G, we can at any time use::
736 # echo 1G > memory.soft_limit_in_bytes
740 reclaiming memory for balancing between memory cgroups
743 It is recommended to set the soft limit always below the hard limit,
746 .. _cgroup-v1-memory-move-charges:
753 Reading memory.move_charge_at_immigrate will always return 0 and writing
754 to it will always return -EINVAL.
756 9. Memory thresholds
759 Memory cgroup implements memory thresholds using the cgroups notification
760 API (see cgroups.txt). It allows to register multiple memory and memsw
763 To register a threshold, an application must:
765 - create an eventfd using eventfd(2);
766 - open memory.usage_in_bytes or memory.memsw.usage_in_bytes;
767 - write string like "<event_fd> <fd of memory.usage_in_bytes> <threshold>" to
770 Application will be notified through eventfd when memory usage crosses
773 It's applicable for root and non-root cgroup.
775 .. _cgroup-v1-memory-oom-control:
782 memory.oom_control file is for OOM notification and other controls.
784 Memory cgroup implements OOM notifier using the cgroup notification
785 API (See cgroups.txt). It allows to register multiple OOM notification
788 To register a notifier, an application must:
790 - create an eventfd using eventfd(2)
791 - open memory.oom_control file
792 - write string like "<event_fd> <fd of memory.oom_control>" to
798 You can disable the OOM-killer by writing "1" to memory.oom_control file, as:
800 #echo 1 > memory.oom_control
802 If OOM-killer is disabled, tasks under cgroup will hang/sleep
803 in memory cgroup's OOM-waitqueue when they request accountable memory.
805 For running them, you have to relax the memory cgroup's OOM status by
809 To reduce usage,
812 * move some tasks to other group with account migration.
819 - oom_kill_disable 0 or 1
820 (if 1, oom-killer is disabled)
821 - under_oom 0 or 1
822 (if 1, the memory cgroup is under OOM, tasks may be stopped.)
823 - oom_kill integer counter
824 The number of processes belonging to this cgroup killed by any
827 11. Memory Pressure (DEPRECATED)
832 The pressure level notifications can be used to monitor the memory
834 different strategies of managing their memory resources. The pressure
837 The "low" level means that the system is reclaiming memory for new
843 The "medium" level means that the system is experiencing medium memory
845 etc. Upon this event applications may decide to further analyze
846 vmstat/zoneinfo/memcg or internal memory usage statistics and free any
847 resources that can be easily reconstructed or re-read from a disk.
850 about to out of memory (OOM) or even the in-kernel OOM killer is on its
851 way to trigger. Applications should do whatever they can to help the
852 system. It might be too late to consult with vmstat or any other
853 statistics, so it's advisable to take an immediate action.
856 events are not pass-through. For example, you have three cgroups: A->B->C. Now
859 notification, i.e. groups A and B will not receive it. This is done to avoid
861 especially bad if we are low on memory or thrashing. Group B, will receive
866 - "default": this is the default behavior specified above. This mode is the
870 - "hierarchy": events always propagate up to the root, similar to the default
873 example, groups A, B, and C will receive notification of memory pressure.
875 - "local": events are pass-through, i.e. they only receive notifications when
876 memory pressure is experienced in the memcg for which the notification is
878 registered for "local" notification and the group experiences memory
884 specified by a comma-delimited string, i.e. "low,hierarchy" specifies
885 hierarchical, pass-through, notification for all ancestor memcgs. Notification
886 that is the default, non pass-through behavior, does not specify a mode.
887 "medium,local" specifies pass-through notification for the medium level.
889 The file memory.pressure_level is only used to setup an eventfd. To
892 - create an eventfd using eventfd(2);
893 - open memory.pressure_level;
894 - write string as "<event_fd> <fd of memory.pressure_level> <level[,mode]>"
895 to cgroup.event_control.
897 Application will be notified through eventfd when memory pressure is at
898 the specific level (or higher). Read/write operations to
899 memory.pressure_level are no implemented.
904 memory limit, sets up a notification in the cgroup and then makes child
907 # cd /sys/fs/cgroup/memory/
910 # cgroup_event_listener memory.pressure_level low,hierarchy &
911 # echo 8000000 > memory.limit_in_bytes
912 # echo 8000000 > memory.memsw.limit_in_bytes
916 (Expect a bunch of notifications, and eventually, the oom-killer will
922 1. Make per-cgroup scanner reclaim not-shared pages first
923 2. Teach controller to account for shared-pages
930 Overall, the memory controller has been a stable controller and has been
936 .. [1] Singh, Balbir. RFC: Memory Controller, http://lwn.net/Articles/206697/
937 .. [2] Singh, Balbir. Memory Controller (RSS Control),
953 10. Singh, Balbir. Memory controller v6 test results,
954 https://lore.kernel.org/r/20070819094658.654.84837.sendpatchset@balbir-laptop
956 .. [11] Singh, Balbir. Memory controller introduction (v6),
957 https://lore.kernel.org/r/20070817084228.26003.12568.sendpatchset@balbir-laptop
958 .. [12] Corbet, Jonathan, Controlling memory use in cgroups,