1 Memory Resource Controller
3 NOTE: The Memory Resource Controller has been generically been referred
4 to as the memory controller in this document. Do not confuse memory controller
5 used here with the memory controller that is used in hardware.
9 a. Enable control of Anonymous, Page Cache (mapped and unmapped) and
10 Swap Cache memory pages.
11 b. The infrastructure allows easy addition of other types of memory to control
12 c. Provides *zero overhead* for non memory controller users
13 d. Provides a double LRU: global memory pressure causes reclaim from the
14 global LRU; a cgroup on hitting a limit, reclaims from the per
17 Benefits and Purpose of the memory controller
19 The memory controller isolates the memory behaviour of a group of tasks
20 from the rest of the system. The article on LWN [12] mentions some probable
21 uses of the memory controller. The memory controller can be used to
23 a. Isolate an application or a group of applications
24 Memory hungry applications can be isolated and limited to a smaller
26 b. Create a cgroup with limited amount of memory, this can be used
27 as a good alternative to booting with mem=XXXX.
28 c. Virtualization solutions can control the amount of memory they want
29 to assign to a virtual machine instance.
30 d. A CD/DVD burner could control the amount of memory used by the
31 rest of the system to ensure that burning does not fail due to lack
33 e. There are several other use cases, find one or use the controller just
34 for fun (to learn and hack on the VM subsystem).
38 The memory controller has a long history. A request for comments for the memory
39 controller was posted by Balbir Singh [1]. At the time the RFC was posted
40 there were several implementations for memory control. The goal of the
41 RFC was to build consensus and agreement for the minimal features required
42 for memory control. The first RSS controller was posted by Balbir Singh[2]
43 in Feb 2007. Pavel Emelianov [3][4][5] has since posted three versions of the
44 RSS controller. At OLS, at the resource management BoF, everyone suggested
45 that we handle both page cache and RSS together. Another request was raised
46 to allow user space handling of OOM. The current memory controller is
47 at version 6; it combines both mapped (RSS) and unmapped Page
52 Memory is a unique resource in the sense that it is present in a limited
53 amount. If a task requires a lot of CPU processing, the task can spread
54 its processing over a period of hours, days, months or years, but with
55 memory, the same physical memory needs to be reused to accomplish the task.
57 The memory controller implementation has been divided into phases. These
61 2. mlock(2) controller
62 3. Kernel user memory accounting and slab control
63 4. user mappings length controller
65 The memory controller is the first controller developed.
69 The core of the design is a counter called the res_counter. The res_counter
70 tracks the current memory usage and limit of the group of processes associated
71 with the controller. Each cgroup has a memory controller specific data
72 structure (mem_cgroup) associated with it.
76 +--------------------+
79 +--------------------+
82 +---------------+ | +---------------+
83 | mm_struct | |.... | mm_struct |
85 +---------------+ | +---------------+
89 +---------------+ +------+--------+
90 | page +----------> page_cgroup|
92 +---------------+ +---------------+
94 (Figure 1: Hierarchy of Accounting)
97 Figure 1 shows the important aspects of the controller
99 1. Accounting happens per cgroup
100 2. Each mm_struct knows about which cgroup it belongs to
101 3. Each page has a pointer to the page_cgroup, which in turn knows the
104 The accounting is done as follows: mem_cgroup_charge() is invoked to setup
105 the necessary data structures and check if the cgroup that is being charged
106 is over its limit. If it is then reclaim is invoked on the cgroup.
107 More details can be found in the reclaim section of this document.
108 If everything goes well, a page meta-data-structure called page_cgroup is
109 allocated and associated with the page. This routine also adds the page to
112 2.2.1 Accounting details
114 All mapped anon pages (RSS) and cache pages (Page Cache) are accounted.
115 (some pages which never be reclaimable and will not be on global LRU
116 are not accounted. we just accounts pages under usual vm management.)
118 RSS pages are accounted at page_fault unless they've already been accounted
119 for earlier. A file page will be accounted for as Page Cache when it's
120 inserted into inode (radix-tree). While it's mapped into the page tables of
121 processes, duplicate accounting is carefully avoided.
123 A RSS page is unaccounted when it's fully unmapped. A PageCache page is
124 unaccounted when it's removed from radix-tree.
126 At page migration, accounting information is kept.
128 Note: we just account pages-on-lru because our purpose is to control amount
129 of used pages. not-on-lru pages are tend to be out-of-control from vm view.
131 2.3 Shared Page Accounting
133 Shared pages are accounted on the basis of the first touch approach. The
134 cgroup that first touches a page is accounted for the page. The principle
135 behind this approach is that a cgroup that aggressively uses a shared
136 page will eventually get charged for it (once it is uncharged from
137 the cgroup that brought it in -- this will happen on memory pressure).
139 Exception: If CONFIG_CGROUP_CGROUP_MEM_RES_CTLR_SWAP is not used..
140 When you do swapoff and make swapped-out pages of shmem(tmpfs) to
141 be backed into memory in force, charges for pages are accounted against the
142 caller of swapoff rather than the users of shmem.
145 2.4 Swap Extension (CONFIG_CGROUP_MEM_RES_CTLR_SWAP)
146 Swap Extension allows you to record charge for swap. A swapped-in page is
147 charged back to original page allocator if possible.
149 When swap is accounted, following files are added.
150 - memory.memsw.usage_in_bytes.
151 - memory.memsw.limit_in_bytes.
153 usage of mem+swap is limited by memsw.limit_in_bytes.
155 * why 'mem+swap' rather than swap.
156 The global LRU(kswapd) can swap out arbitrary pages. Swap-out means
157 to move account from memory to swap...there is no change in usage of
158 mem+swap. In other words, when we want to limit the usage of swap without
159 affecting global LRU, mem+swap limit is better than just limiting swap from
162 * What happens when a cgroup hits memory.memsw.limit_in_bytes
163 When a cgroup his memory.memsw.limit_in_bytes, it's useless to do swap-out
164 in this cgroup. Then, swap-out will not be done by cgroup routine and file
165 caches are dropped. But as mentioned above, global LRU can do swapout memory
166 from it for sanity of the system's memory management state. You can't forbid
171 Each cgroup maintains a per cgroup LRU that consists of an active
172 and inactive list. When a cgroup goes over its limit, we first try
173 to reclaim memory from the cgroup so as to make space for the new
174 pages that the cgroup has touched. If the reclaim is unsuccessful,
175 an OOM routine is invoked to select and kill the bulkiest task in the
178 The reclaim algorithm has not been modified for cgroups, except that
179 pages that are selected for reclaiming come from the per cgroup LRU
182 NOTE: Reclaim does not work for the root cgroup, since we cannot set any
183 limits on the root cgroup.
187 The memory controller uses the following hierarchy
189 1. zone->lru_lock is used for selecting pages to be isolated
190 2. mem->per_zone->lru_lock protects the per cgroup LRU (per zone)
191 3. lock_page_cgroup() is used to protect page->page_cgroup
197 a. Enable CONFIG_CGROUPS
198 b. Enable CONFIG_RESOURCE_COUNTERS
199 c. Enable CONFIG_CGROUP_MEM_RES_CTLR
201 1. Prepare the cgroups
203 # mount -t cgroup none /cgroups -o memory
205 2. Make the new group and move bash into it
207 # echo $$ > /cgroups/0/tasks
209 Since now we're in the 0 cgroup,
210 We can alter the memory limit:
211 # echo 4M > /cgroups/0/memory.limit_in_bytes
213 NOTE: We can use a suffix (k, K, m, M, g or G) to indicate values in kilo,
215 NOTE: We can write "-1" to reset the *.limit_in_bytes(unlimited).
216 NOTE: We cannot set limits on the root cgroup any more.
218 # cat /cgroups/0/memory.limit_in_bytes
221 NOTE: The interface has now changed to display the usage in bytes
224 We can check the usage:
225 # cat /cgroups/0/memory.usage_in_bytes
228 A successful write to this file does not guarantee a successful set of
229 this limit to the value written into the file. This can be due to a
230 number of factors, such as rounding up to page boundaries or the total
231 availability of memory on the system. The user is required to re-read
232 this file after a write to guarantee the value committed by the kernel.
234 # echo 1 > memory.limit_in_bytes
235 # cat memory.limit_in_bytes
238 The memory.failcnt field gives the number of times that the cgroup limit was
241 The memory.stat file gives accounting information. Now, the number of
242 caches, RSS and Active pages/Inactive pages are shown.
246 Balbir posted lmbench, AIM9, LTP and vmmstress results [10] and [11].
247 Apart from that v6 has been tested with several applications and regular
248 daily use. The controller has also been tested on the PPC64, x86_64 and
253 Sometimes a user might find that the application under a cgroup is
254 terminated. There are several causes for this:
256 1. The cgroup limit is too low (just too low to do anything useful)
257 2. The user is using anonymous memory and swap is turned off or too low
259 A sync followed by echo 1 > /proc/sys/vm/drop_caches will help get rid of
260 some of the pages cached in the cgroup (page cache pages).
264 When a task migrates from one cgroup to another, it's charge is not
265 carried forward by default. The pages allocated from the original cgroup still
266 remain charged to it, the charge is dropped when the page is freed or
269 Note: You can move charges of a task along with task migration. See 8.
271 4.3 Removing a cgroup
273 A cgroup can be removed by rmdir, but as discussed in sections 4.1 and 4.2, a
274 cgroup might have some charge associated with it, even though all
275 tasks have migrated away from it.
276 Such charges are freed(at default) or moved to its parent. When moved,
277 both of RSS and CACHES are moved to parent.
278 If both of them are busy, rmdir() returns -EBUSY. See 5.1 Also.
280 Charges recorded in swap information is not updated at removal of cgroup.
281 Recorded information is discarded and a cgroup which uses swap (swapcache)
282 will be charged as a new owner of it.
288 memory.force_empty interface is provided to make cgroup's memory usage empty.
289 You can use this interface only when the cgroup has no tasks.
290 When writing anything to this
292 # echo 0 > memory.force_empty
294 Almost all pages tracked by this memcg will be unmapped and freed. Some of
295 pages cannot be freed because it's locked or in-use. Such pages are moved
296 to parent and this cgroup will be empty. But this may return -EBUSY in
299 Typical use case of this interface is that calling this before rmdir().
300 Because rmdir() moves all pages to parent, some out-of-use page caches can be
301 moved to the parent. If you want to avoid that, force_empty will be useful.
305 memory.stat file includes following statistics
307 cache - # of bytes of page cache memory.
308 rss - # of bytes of anonymous and swap cache memory.
309 pgpgin - # of pages paged in (equivalent to # of charging events).
310 pgpgout - # of pages paged out (equivalent to # of uncharging events).
311 active_anon - # of bytes of anonymous and swap cache memory on active
313 inactive_anon - # of bytes of anonymous memory and swap cache memory on
315 active_file - # of bytes of file-backed memory on active lru list.
316 inactive_file - # of bytes of file-backed memory on inactive lru list.
317 unevictable - # of bytes of memory that cannot be reclaimed (mlocked etc).
319 The following additional stats are dependent on CONFIG_DEBUG_VM.
321 inactive_ratio - VM internal parameter. (see mm/page_alloc.c)
322 recent_rotated_anon - VM internal parameter. (see mm/vmscan.c)
323 recent_rotated_file - VM internal parameter. (see mm/vmscan.c)
324 recent_scanned_anon - VM internal parameter. (see mm/vmscan.c)
325 recent_scanned_file - VM internal parameter. (see mm/vmscan.c)
328 recent_rotated means recent frequency of lru rotation.
329 recent_scanned means recent # of scans to lru.
330 showing for better debug please see the code for meanings.
333 Only anonymous and swap cache memory is listed as part of 'rss' stat.
334 This should not be confused with the true 'resident set size' or the
335 amount of physical memory used by the cgroup. Per-cgroup rss
336 accounting is not done yet.
339 Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only.
341 Following cgroups' swapiness can't be changed.
342 - root cgroup (uses /proc/sys/vm/swappiness).
343 - a cgroup which uses hierarchy and it has child cgroup.
344 - a cgroup which uses hierarchy and not the root of hierarchy.
349 The memory controller supports a deep hierarchy and hierarchical accounting.
350 The hierarchy is created by creating the appropriate cgroups in the
351 cgroup filesystem. Consider for example, the following cgroup filesystem
362 In the diagram above, with hierarchical accounting enabled, all memory
363 usage of e, is accounted to its ancestors up until the root (i.e, c and root),
364 that has memory.use_hierarchy enabled. If one of the ancestors goes over its
365 limit, the reclaim algorithm reclaims from the tasks in the ancestor and the
366 children of the ancestor.
368 6.1 Enabling hierarchical accounting and reclaim
370 The memory controller by default disables the hierarchy feature. Support
371 can be enabled by writing 1 to memory.use_hierarchy file of the root cgroup
373 # echo 1 > memory.use_hierarchy
375 The feature can be disabled by
377 # echo 0 > memory.use_hierarchy
379 NOTE1: Enabling/disabling will fail if the cgroup already has other
380 cgroups created below it.
382 NOTE2: This feature can be enabled/disabled per subtree.
386 Soft limits allow for greater sharing of memory. The idea behind soft limits
387 is to allow control groups to use as much of the memory as needed, provided
389 a. There is no memory contention
390 b. They do not exceed their hard limit
392 When the system detects memory contention or low memory control groups
393 are pushed back to their soft limits. If the soft limit of each control
394 group is very high, they are pushed back as much as possible to make
395 sure that one control group does not starve the others of memory.
397 Please note that soft limits is a best effort feature, it comes with
398 no guarantees, but it does its best to make sure that when memory is
399 heavily contended for, memory is allocated based on the soft limit
400 hints/setup. Currently soft limit based reclaim is setup such that
401 it gets invoked from balance_pgdat (kswapd).
405 Soft limits can be setup by using the following commands (in this example we
406 assume a soft limit of 256 megabytes)
408 # echo 256M > memory.soft_limit_in_bytes
410 If we want to change this to 1G, we can at any time use
412 # echo 1G > memory.soft_limit_in_bytes
414 NOTE1: Soft limits take effect over a long period of time, since they involve
415 reclaiming memory for balancing between memory cgroups
416 NOTE2: It is recommended to set the soft limit always below the hard limit,
417 otherwise the hard limit will take precedence.
419 8. Move charges at task migration
421 Users can move charges associated with a task along with task migration, that
422 is, uncharge task's pages from the old cgroup and charge them to the new cgroup.
426 This feature is disabled by default. It can be enabled(and disabled again) by
427 writing to memory.move_charge_at_immigrate of the destination cgroup.
429 If you want to enable it:
431 # echo (some positive value) > memory.move_charge_at_immigrate
433 Note: Each bits of move_charge_at_immigrate has its own meaning about what type
434 of charges should be moved. See 8.2 for details.
435 Note: Charges are moved only when you move mm->owner, IOW, a leader of a thread
437 Note: If we cannot find enough space for the task in the destination cgroup, we
438 try to make space by reclaiming memory. Task migration may fail if we
439 cannot make enough space.
440 Note: It can take several seconds if you move charges in giga bytes order.
442 And if you want disable it again:
444 # echo 0 > memory.move_charge_at_immigrate
446 8.2 Type of charges which can be move
448 Each bits of move_charge_at_immigrate has its own meaning about what type of
449 charges should be moved.
451 bit | what type of charges would be moved ?
452 -----+------------------------------------------------------------------------
453 0 | A charge of an anonymous page(or swap of it) used by the target task.
454 | Those pages and swaps must be used only by the target task. You must
455 | enable Swap Extension(see 2.4) to enable move of swap charges.
457 Note: Those pages and swaps must be charged to the old cgroup.
458 Note: More type of pages(e.g. file cache, shmem,) will be supported by other
463 - Add support for other types of pages(e.g. file cache, shmem, etc.).
464 - Implement madvise(2) to let users decide the vma to be moved or not to be
466 - All of moving charge operations are done under cgroup_mutex. It's not good
467 behavior to hold the mutex too long, so we may need some trick.
471 1. Add support for accounting huge pages (as a separate controller)
472 2. Make per-cgroup scanner reclaim not-shared pages first
473 3. Teach controller to account for shared-pages
474 4. Start reclamation in the background when the limit is
475 not yet hit but the usage is getting closer
479 Overall, the memory controller has been a stable controller and has been
480 commented and discussed quite extensively in the community.
484 1. Singh, Balbir. RFC: Memory Controller, http://lwn.net/Articles/206697/
485 2. Singh, Balbir. Memory Controller (RSS Control),
486 http://lwn.net/Articles/222762/
487 3. Emelianov, Pavel. Resource controllers based on process cgroups
488 http://lkml.org/lkml/2007/3/6/198
489 4. Emelianov, Pavel. RSS controller based on process cgroups (v2)
490 http://lkml.org/lkml/2007/4/9/78
491 5. Emelianov, Pavel. RSS controller based on process cgroups (v3)
492 http://lkml.org/lkml/2007/5/30/244
493 6. Menage, Paul. Control Groups v10, http://lwn.net/Articles/236032/
494 7. Vaidyanathan, Srinivasan, Control Groups: Pagecache accounting and control
495 subsystem (v3), http://lwn.net/Articles/235534/
496 8. Singh, Balbir. RSS controller v2 test results (lmbench),
497 http://lkml.org/lkml/2007/5/17/232
498 9. Singh, Balbir. RSS controller v2 AIM9 results
499 http://lkml.org/lkml/2007/5/18/1
500 10. Singh, Balbir. Memory controller v6 test results,
501 http://lkml.org/lkml/2007/8/19/36
502 11. Singh, Balbir. Memory controller introduction (v6),
503 http://lkml.org/lkml/2007/8/17/69
504 12. Corbet, Jonathan, Controlling memory use in cgroups,
505 http://lwn.net/Articles/243795/