memory.txt 35.7 KB
Newer Older
1 2
Memory Resource Controller

3 4 5
NOTE: The Memory Resource Controller has generically been referred to as the
      memory controller in this document. Do not confuse memory controller
      used here with the memory controller that is used in hardware.
B
Balbir Singh 已提交
6

K
KAMEZAWA Hiroyuki 已提交
7 8 9 10 11 12
(For editors)
In this document:
      When we mention a cgroup (cgroupfs's directory) with memory controller,
      we call it "memory cgroup". When you see git-log and source code, you'll
      see patch's title and function names tend to use "memcg".
      In this document, we avoid using it.
B
Balbir Singh 已提交
13 14 15 16 17 18 19 20

Benefits and Purpose of the memory controller

The memory controller isolates the memory behaviour of a group of tasks
from the rest of the system. The article on LWN [12] mentions some probable
uses of the memory controller. The memory controller can be used to

a. Isolate an application or a group of applications
21
   Memory-hungry applications can be isolated and limited to a smaller
B
Balbir Singh 已提交
22
   amount of memory.
23
b. Create a cgroup with a limited amount of memory; this can be used
B
Balbir Singh 已提交
24 25 26 27 28 29
   as a good alternative to booting with mem=XXXX.
c. Virtualization solutions can control the amount of memory they want
   to assign to a virtual machine instance.
d. A CD/DVD burner could control the amount of memory used by the
   rest of the system to ensure that burning does not fail due to lack
   of available memory.
30
e. There are several other use cases; find one or use the controller just
B
Balbir Singh 已提交
31 32
   for fun (to learn and hack on the VM subsystem).

K
KAMEZAWA Hiroyuki 已提交
33 34 35 36
Current Status: linux-2.6.34-mmotm(development version of 2010/April)

Features:
 - accounting anonymous pages, file caches, swap caches usage and limiting them.
37
 - pages are linked to per-memcg LRU exclusively, and there is no global LRU.
K
KAMEZAWA Hiroyuki 已提交
38 39 40
 - optionally, memory+swap usage can be accounted and limited.
 - hierarchical accounting
 - soft limit
41
 - moving (recharging) account at moving a task is selectable.
K
KAMEZAWA Hiroyuki 已提交
42
 - usage threshold notifier
43
 - memory pressure notifier
K
KAMEZAWA Hiroyuki 已提交
44 45 46
 - oom-killer disable knob and oom-notifier
 - Root cgroup has no limit controls.

47
 Kernel memory support is a work in progress, and the current version provides
48
 basically functionality. (See Section 2.7)
K
KAMEZAWA Hiroyuki 已提交
49 50 51 52 53 54

Brief summary of control files.

 tasks				 # attach a task(thread) and show list of threads
 cgroup.procs			 # show list of processes
 cgroup.event_control		 # an interface for event_fd()
55 56 57 58
 memory.usage_in_bytes		 # show current res_counter usage for memory
				 (See 5.5 for details)
 memory.memsw.usage_in_bytes	 # show current res_counter usage for memory+Swap
				 (See 5.5 for details)
K
KAMEZAWA Hiroyuki 已提交
59 60 61 62 63
 memory.limit_in_bytes		 # set/show limit of memory usage
 memory.memsw.limit_in_bytes	 # set/show limit of memory+Swap usage
 memory.failcnt			 # show the number of memory usage hits limits
 memory.memsw.failcnt		 # show the number of memory+Swap hits limits
 memory.max_usage_in_bytes	 # show max memory usage recorded
64
 memory.memsw.max_usage_in_bytes # show max memory+Swap usage recorded
K
KAMEZAWA Hiroyuki 已提交
65 66 67 68
 memory.soft_limit_in_bytes	 # set/show soft limit of memory usage
 memory.stat			 # show various statistics
 memory.use_hierarchy		 # set/show hierarchical account enabled
 memory.force_empty		 # trigger forced move charge to parent
69
 memory.pressure_level		 # set memory pressure notifications
K
KAMEZAWA Hiroyuki 已提交
70 71 72 73
 memory.swappiness		 # set/show swappiness parameter of vmscan
				 (See sysctl's vm.swappiness)
 memory.move_charge_at_immigrate # set/show controls of moving charges
 memory.oom_control		 # set/show oom controls.
74
 memory.numa_stat		 # show the number of memory usage per numa node
K
KAMEZAWA Hiroyuki 已提交
75

76 77 78 79 80
 memory.kmem.limit_in_bytes      # set/show hard limit for kernel memory
 memory.kmem.usage_in_bytes      # show current kernel memory allocation
 memory.kmem.failcnt             # show the number of kernel memory usage hits limits
 memory.kmem.max_usage_in_bytes  # show max kernel memory usage recorded

81
 memory.kmem.tcp.limit_in_bytes  # set/show hard limit for tcp buf memory
82
 memory.kmem.tcp.usage_in_bytes  # show current tcp buf memory allocation
83 84
 memory.kmem.tcp.failcnt            # show the number of tcp buf memory usage hits limits
 memory.kmem.tcp.max_usage_in_bytes # show max tcp buf memory usage recorded
85

B
Balbir Singh 已提交
86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
1. History

The memory controller has a long history. A request for comments for the memory
controller was posted by Balbir Singh [1]. At the time the RFC was posted
there were several implementations for memory control. The goal of the
RFC was to build consensus and agreement for the minimal features required
for memory control. The first RSS controller was posted by Balbir Singh[2]
in Feb 2007. Pavel Emelianov [3][4][5] has since posted three versions of the
RSS controller. At OLS, at the resource management BoF, everyone suggested
that we handle both page cache and RSS together. Another request was raised
to allow user space handling of OOM. The current memory controller is
at version 6; it combines both mapped (RSS) and unmapped Page
Cache Control [11].

2. Memory Control

Memory is a unique resource in the sense that it is present in a limited
amount. If a task requires a lot of CPU processing, the task can spread
its processing over a period of hours, days, months or years, but with
memory, the same physical memory needs to be reused to accomplish the task.

The memory controller implementation has been divided into phases. These
are:

1. Memory controller
2. mlock(2) controller
3. Kernel user memory accounting and slab control
4. user mappings length controller

The memory controller is the first controller developed.

2.1. Design

The core of the design is a counter called the res_counter. The res_counter
tracks the current memory usage and limit of the group of processes associated
with the controller. Each cgroup has a memory controller specific data
structure (mem_cgroup) associated with it.

2.2. Accounting

		+--------------------+
		|  mem_cgroup     |
		|  (res_counter)     |
		+--------------------+
		 /            ^      \
		/             |       \
           +---------------+  |        +---------------+
           | mm_struct     |  |....    | mm_struct     |
           |               |  |        |               |
           +---------------+  |        +---------------+
                              |
                              + --------------+
                                              |
           +---------------+           +------+--------+
           | page          +---------->  page_cgroup|
           |               |           |               |
           +---------------+           +---------------+

             (Figure 1: Hierarchy of Accounting)


Figure 1 shows the important aspects of the controller

1. Accounting happens per cgroup
2. Each mm_struct knows about which cgroup it belongs to
3. Each page has a pointer to the page_cgroup, which in turn knows the
   cgroup it belongs to

154 155 156
The accounting is done as follows: mem_cgroup_charge_common() is invoked to
set up the necessary data structures and check if the cgroup that is being
charged is over its limit. If it is, then reclaim is invoked on the cgroup.
B
Balbir Singh 已提交
157 158
More details can be found in the reclaim section of this document.
If everything goes well, a page meta-data-structure called page_cgroup is
K
KAMEZAWA Hiroyuki 已提交
159 160
updated. page_cgroup has its own LRU on cgroup.
(*) page_cgroup structure is allocated at boot/memory-hotplug time.
B
Balbir Singh 已提交
161 162 163

2.2.1 Accounting details

164
All mapped anon pages (RSS) and cache pages (Page Cache) are accounted.
165
Some pages which are never reclaimable and will not be on the LRU
K
KAMEZAWA Hiroyuki 已提交
166
are not accounted. We just account pages under usual VM management.
167 168 169 170 171 172

RSS pages are accounted at page_fault unless they've already been accounted
for earlier. A file page will be accounted for as Page Cache when it's
inserted into inode (radix-tree). While it's mapped into the page tables of
processes, duplicate accounting is carefully avoided.

173
An RSS page is unaccounted when it's fully unmapped. A PageCache page is
K
KAMEZAWA Hiroyuki 已提交
174 175
unaccounted when it's removed from radix-tree. Even if RSS pages are fully
unmapped (by kswapd), they may exist as SwapCache in the system until they
176
are really freed. Such SwapCaches are also accounted.
K
KAMEZAWA Hiroyuki 已提交
177 178
A swapped-in page is not accounted until it's mapped.

179
Note: The kernel does swapin-readahead and reads multiple swaps at once.
K
KAMEZAWA Hiroyuki 已提交
180 181
This means swapped-in pages may contain pages for other tasks than a task
causing page fault. So, we avoid accounting at swap-in I/O.
182 183 184

At page migration, accounting information is kept.

K
KAMEZAWA Hiroyuki 已提交
185 186
Note: we just account pages-on-LRU because our purpose is to control amount
of used pages; not-on-LRU pages tend to be out-of-control from VM view.
B
Balbir Singh 已提交
187 188 189 190 191 192 193 194 195

2.3 Shared Page Accounting

Shared pages are accounted on the basis of the first touch approach. The
cgroup that first touches a page is accounted for the page. The principle
behind this approach is that a cgroup that aggressively uses a shared
page will eventually get charged for it (once it is uncharged from
the cgroup that brought it in -- this will happen on memory pressure).

196 197 198
But see section 8.2: when moving a task to another cgroup, its pages may
be recharged to the new cgroup, if move_charge_at_immigrate has been chosen.

199
Exception: If CONFIG_MEMCG_SWAP is not used.
200
When you do swapoff and make swapped-out pages of shmem(tmpfs) to
K
KAMEZAWA Hiroyuki 已提交
201 202 203
be backed into memory in force, charges for pages are accounted against the
caller of swapoff rather than the users of shmem.

A
Andrew Morton 已提交
204
2.4 Swap Extension (CONFIG_MEMCG_SWAP)
K
KAMEZAWA Hiroyuki 已提交
205

206 207 208 209 210 211 212
Swap Extension allows you to record charge for swap. A swapped-in page is
charged back to original page allocator if possible.

When swap is accounted, following files are added.
 - memory.memsw.usage_in_bytes.
 - memory.memsw.limit_in_bytes.

K
KAMEZAWA Hiroyuki 已提交
213 214 215 216 217 218
memsw means memory+swap. Usage of memory+swap is limited by
memsw.limit_in_bytes.

Example: Assume a system with 4G of swap. A task which allocates 6G of memory
(by mistake) under 2G memory limitation will use all swap.
In this case, setting memsw.limit_in_bytes=3G will prevent bad use of swap.
219
By using the memsw limit, you can avoid system OOM which can be caused by swap
K
KAMEZAWA Hiroyuki 已提交
220
shortage.
221

K
KAMEZAWA Hiroyuki 已提交
222
* why 'memory+swap' rather than swap.
223 224
The global LRU(kswapd) can swap out arbitrary pages. Swap-out means
to move account from memory to swap...there is no change in usage of
K
KAMEZAWA Hiroyuki 已提交
225 226
memory+swap. In other words, when we want to limit the usage of swap without
affecting global LRU, memory+swap limit is better than just limiting swap from
227
an OS point of view.
228 229

* What happens when a cgroup hits memory.memsw.limit_in_bytes
230
When a cgroup hits memory.memsw.limit_in_bytes, it's useless to do swap-out
231 232 233 234
in this cgroup. Then, swap-out will not be done by cgroup routine and file
caches are dropped. But as mentioned above, global LRU can do swapout memory
from it for sanity of the system's memory management state. You can't forbid
it by cgroup.
235 236

2.5 Reclaim
B
Balbir Singh 已提交
237

K
KAMEZAWA Hiroyuki 已提交
238 239
Each cgroup maintains a per cgroup LRU which has the same structure as
global VM. When a cgroup goes over its limit, we first try
B
Balbir Singh 已提交
240 241 242
to reclaim memory from the cgroup so as to make space for the new
pages that the cgroup has touched. If the reclaim is unsuccessful,
an OOM routine is invoked to select and kill the bulkiest task in the
K
KAMEZAWA Hiroyuki 已提交
243
cgroup. (See 10. OOM Control below.)
B
Balbir Singh 已提交
244 245

The reclaim algorithm has not been modified for cgroups, except that
246
pages that are selected for reclaiming come from the per-cgroup LRU
B
Balbir Singh 已提交
247 248
list.

249 250 251
NOTE: Reclaim does not work for the root cgroup, since we cannot set any
limits on the root cgroup.

252 253
Note2: When panic_on_oom is set to "2", the whole system will panic.

K
KAMEZAWA Hiroyuki 已提交
254 255 256
When oom event notifier is registered, event will be delivered.
(See oom_control section)

K
KAMEZAWA Hiroyuki 已提交
257
2.6 Locking
B
Balbir Singh 已提交
258

K
KAMEZAWA Hiroyuki 已提交
259 260
   lock_page_cgroup()/unlock_page_cgroup() should not be called under
   mapping->tree_lock.
B
Balbir Singh 已提交
261

K
KAMEZAWA Hiroyuki 已提交
262 263 264 265 266 267 268 269
   Other lock order is following:
   PG_locked.
   mm->page_table_lock
       zone->lru_lock
	  lock_page_cgroup.
  In many cases, just lock_page_cgroup() is called.
  per-zone-per-cgroup LRU (cgroup's private LRU) is just guarded by
  zone->lru_lock, it has no lock of its own.
B
Balbir Singh 已提交
270

A
Andrew Morton 已提交
271
2.7 Kernel Memory Extension (CONFIG_MEMCG_KMEM)
272 273 274 275 276 277

With the Kernel memory extension, the Memory Controller is able to limit
the amount of kernel memory used by the system. Kernel memory is fundamentally
different than user memory, since it can't be swapped out, which makes it
possible to DoS the system by consuming too much of this precious resource.

278 279 280 281 282 283 284 285 286 287 288 289
Kernel memory won't be accounted at all until limit on a group is set. This
allows for existing setups to continue working without disruption.  The limit
cannot be set if the cgroup have children, or if there are already tasks in the
cgroup. Attempting to set the limit under those conditions will return -EBUSY.
When use_hierarchy == 1 and a group is accounted, its children will
automatically be accounted regardless of their limit value.

After a group is first limited, it will be kept being accounted until it
is removed. The memory limitation itself, can of course be removed by writing
-1 to memory.kmem.limit_in_bytes. In this case, kmem will be accounted, but not
limited.

290
Kernel memory limits are not imposed for the root cgroup. Usage for the root
291 292 293 294 295
cgroup may or may not be accounted. The memory used is accumulated into
memory.kmem.usage_in_bytes, or in a separate counter when it makes sense.
(currently only for tcp).
The main "kmem" counter is fed into the main counter, so kmem charges will
also be visible from the user counter.
296 297 298 299 300 301

Currently no soft limit is implemented for kernel memory. It is future work
to trigger slab reclaim when those limits are reached.

2.7.1 Current Kernel Memory resources accounted

302 303 304 305
* stack pages: every process consumes some stack pages. By accounting into
kernel memory, we prevent new processes from being created when the kernel
memory usage is too high.

306 307 308 309 310 311 312
* slab pages: pages allocated by the SLAB or SLUB allocator are tracked. A copy
of each kmem_cache is created everytime the cache is touched by the first time
from inside the memcg. The creation is done lazily, so some objects can still be
skipped while the cache is being created. All objects in a slab page should
belong to the same memcg. This only fails to hold when a task is migrated to a
different memcg during the page allocation by the cache.

G
Glauber Costa 已提交
313 314 315
* sockets memory pressure: some sockets protocols have memory pressure
thresholds. The Memory Controller allows them to be controlled individually
per cgroup, instead of globally.
316

G
Glauber Costa 已提交
317 318
* tcp memory pressure: sockets memory pressure for the tcp protocol.

319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344
2.7.3 Common use cases

Because the "kmem" counter is fed to the main user counter, kernel memory can
never be limited completely independently of user memory. Say "U" is the user
limit, and "K" the kernel limit. There are three possible ways limits can be
set:

    U != 0, K = unlimited:
    This is the standard memcg limitation mechanism already present before kmem
    accounting. Kernel memory is completely ignored.

    U != 0, K < U:
    Kernel memory is a subset of the user memory. This setup is useful in
    deployments where the total amount of memory per-cgroup is overcommited.
    Overcommiting kernel memory limits is definitely not recommended, since the
    box can still run out of non-reclaimable memory.
    In this case, the admin could set up K so that the sum of all groups is
    never greater than the total memory, and freely set U at the cost of his
    QoS.

    U != 0, K >= U:
    Since kmem charges will also be fed to the user counter and reclaim will be
    triggered for the cgroup for both kinds of memory. This setup gives the
    admin a unified view of memory, and it is also useful for people who just
    want to track kernel memory usage.

B
Balbir Singh 已提交
345 346 347 348 349 350
3. User Interface

0. Configuration

a. Enable CONFIG_CGROUPS
b. Enable CONFIG_RESOURCE_COUNTERS
A
Andrew Morton 已提交
351 352
c. Enable CONFIG_MEMCG
d. Enable CONFIG_MEMCG_SWAP (to use swap extension)
353
d. Enable CONFIG_MEMCG_KMEM (to use kmem extension)
B
Balbir Singh 已提交
354

355 356 357 358
1. Prepare the cgroups (see cgroups.txt, Why are cgroups needed?)
# mount -t tmpfs none /sys/fs/cgroup
# mkdir /sys/fs/cgroup/memory
# mount -t cgroup none /sys/fs/cgroup/memory -o memory
B
Balbir Singh 已提交
359 360

2. Make the new group and move bash into it
361 362
# mkdir /sys/fs/cgroup/memory/0
# echo $$ > /sys/fs/cgroup/memory/0/tasks
B
Balbir Singh 已提交
363

K
KAMEZAWA Hiroyuki 已提交
364
Since now we're in the 0 cgroup, we can alter the memory limit:
365
# echo 4M > /sys/fs/cgroup/memory/0/memory.limit_in_bytes
366 367

NOTE: We can use a suffix (k, K, m, M, g or G) to indicate values in kilo,
K
KAMEZAWA Hiroyuki 已提交
368 369
mega or gigabytes. (Here, Kilo, Mega, Giga are Kibibytes, Mebibytes, Gibibytes.)

370
NOTE: We can write "-1" to reset the *.limit_in_bytes(unlimited).
371
NOTE: We cannot set limits on the root cgroup any more.
372

373
# cat /sys/fs/cgroup/memory/0/memory.limit_in_bytes
374
4194304
375

B
Balbir Singh 已提交
376
We can check the usage:
377
# cat /sys/fs/cgroup/memory/0/memory.usage_in_bytes
378
1216512
379

380
A successful write to this file does not guarantee a successful setting of
K
KAMEZAWA Hiroyuki 已提交
381
this limit to the value written into the file. This can be due to a
382
number of factors, such as rounding up to page boundaries or the total
K
KAMEZAWA Hiroyuki 已提交
383
availability of memory on the system. The user is required to re-read
384 385
this file after a write to guarantee the value committed by the kernel.

386
# echo 1 > memory.limit_in_bytes
387
# cat memory.limit_in_bytes
388
4096
B
Balbir Singh 已提交
389 390 391 392

The memory.failcnt field gives the number of times that the cgroup limit was
exceeded.

393 394 395
The memory.stat file gives accounting information. Now, the number of
caches, RSS and Active pages/Inactive pages are shown.

B
Balbir Singh 已提交
396 397
4. Testing

K
KAMEZAWA Hiroyuki 已提交
398 399 400 401 402 403 404 405 406 407 408 409
For testing features and implementation, see memcg_test.txt.

Performance test is also important. To see pure memory controller's overhead,
testing on tmpfs will give you good numbers of small overheads.
Example: do kernel make on tmpfs.

Page-fault scalability is also important. At measuring parallel
page fault test, multi-process test may be better than multi-thread
test because it has noise of shared objects/status.

But the above two are testing extreme situations.
Trying usual test under memory controller is always helpful.
B
Balbir Singh 已提交
410 411 412 413

4.1 Troubleshooting

Sometimes a user might find that the application under a cgroup is
414
terminated by the OOM killer. There are several causes for this:
B
Balbir Singh 已提交
415 416 417 418 419 420 421

1. The cgroup limit is too low (just too low to do anything useful)
2. The user is using anonymous memory and swap is turned off or too low

A sync followed by echo 1 > /proc/sys/vm/drop_caches will help get rid of
some of the pages cached in the cgroup (page cache pages).

422
To know what happens, disabling OOM_Kill as per "10. OOM Control" (below) and
K
KAMEZAWA Hiroyuki 已提交
423 424
seeing what happens will be helpful.

B
Balbir Singh 已提交
425 426
4.2 Task migration

427
When a task migrates from one cgroup to another, its charge is not
428
carried forward by default. The pages allocated from the original cgroup still
B
Balbir Singh 已提交
429 430 431
remain charged to it, the charge is dropped when the page is freed or
reclaimed.

K
KAMEZAWA Hiroyuki 已提交
432 433
You can move charges of a task along with task migration.
See 8. "Move charges at task migration"
434

B
Balbir Singh 已提交
435 436 437 438
4.3 Removing a cgroup

A cgroup can be removed by rmdir, but as discussed in sections 4.1 and 4.2, a
cgroup might have some charge associated with it, even though all
K
KAMEZAWA Hiroyuki 已提交
439 440 441
tasks have migrated away from it. (because we charge against pages, not
against tasks.)

442 443 444
We move the stats to root (if use_hierarchy==0) or parent (if
use_hierarchy==1), and no change on the charge except uncharging
from the child.
B
Balbir Singh 已提交
445

446 447 448 449
Charges recorded in swap information is not updated at removal of cgroup.
Recorded information is discarded and a cgroup which uses swap (swapcache)
will be charged as a new owner of it.

450
About use_hierarchy, see Section 6.
451

452 453 454 455 456 457 458 459 460
5. Misc. interfaces.

5.1 force_empty
  memory.force_empty interface is provided to make cgroup's memory usage empty.
  You can use this interface only when the cgroup has no tasks.
  When writing anything to this

  # echo 0 > memory.force_empty

K
KAMEZAWA Hiroyuki 已提交
461 462
  Almost all pages tracked by this memory cgroup will be unmapped and freed.
  Some pages cannot be freed because they are locked or in-use. Such pages are
463
  moved to parent (if use_hierarchy==1) or root (if use_hierarchy==0) and this
464
  cgroup will be empty.
465

466
  The typical use case for this interface is before calling rmdir().
467 468 469
  Because rmdir() moves all pages to parent, some out-of-use page caches can be
  moved to the parent. If you want to avoid that, force_empty will be useful.

470 471 472 473 474
  Also, note that when memory.kmem.limit_in_bytes is set the charges due to
  kernel pages will still be seen. This is not considered a failure and the
  write will still return success. In this case, it is expected that
  memory.kmem.usage_in_bytes == memory.usage_in_bytes.

475 476
  About use_hierarchy, see Section 6.

K
KOSAKI Motohiro 已提交
477
5.2 stat file
B
Bharata B Rao 已提交
478

479
memory.stat file includes following statistics
B
Bharata B Rao 已提交
480

K
KAMEZAWA Hiroyuki 已提交
481
# per-memory cgroup local status
B
Bharata B Rao 已提交
482
cache		- # of bytes of page cache memory.
483 484 485
rss		- # of bytes of anonymous and swap cache memory (includes
		transparent hugepages).
rss_huge	- # of bytes of anonymous transparent hugepages.
K
KAMEZAWA Hiroyuki 已提交
486
mapped_file	- # of bytes of mapped file (includes tmpfs/shmem)
487 488 489 490 491
pgpgin		- # of charging events to the memory cgroup. The charging
		event happens each time a page is accounted as either mapped
		anon page(RSS) or cache page(Page Cache) to the cgroup.
pgpgout		- # of uncharging events to the memory cgroup. The uncharging
		event happens each time a page is unaccounted from the cgroup.
K
KAMEZAWA Hiroyuki 已提交
492
swap		- # of bytes of swap usage
B
Bharata B Rao 已提交
493
inactive_anon	- # of bytes of anonymous memory and swap cache memory on
K
KAMEZAWA Hiroyuki 已提交
494 495 496 497 498
		LRU list.
active_anon	- # of bytes of anonymous and swap cache memory on active
		inactive LRU list.
inactive_file	- # of bytes of file-backed memory on inactive LRU list.
active_file	- # of bytes of file-backed memory on active LRU list.
B
Bharata B Rao 已提交
499 500
unevictable	- # of bytes of memory that cannot be reclaimed (mlocked etc).

K
KAMEZAWA Hiroyuki 已提交
501 502 503 504 505 506 507
# status considering hierarchy (see memory.use_hierarchy settings)

hierarchical_memory_limit - # of bytes of memory limit with regard to hierarchy
			under which the memory cgroup is
hierarchical_memsw_limit - # of bytes of memory+swap limit with regard to
			hierarchy under which memory cgroup is.

508 509 510 511
total_<counter>		- # hierarchical version of <counter>, which in
			addition to the cgroup's own value includes the
			sum of all hierarchical children's values of
			<counter>, i.e. total_cache
K
KAMEZAWA Hiroyuki 已提交
512 513

# The following additional stats are dependent on CONFIG_DEBUG_VM.
B
Bharata B Rao 已提交
514 515 516 517 518 519 520

recent_rotated_anon	- VM internal parameter. (see mm/vmscan.c)
recent_rotated_file	- VM internal parameter. (see mm/vmscan.c)
recent_scanned_anon	- VM internal parameter. (see mm/vmscan.c)
recent_scanned_file	- VM internal parameter. (see mm/vmscan.c)

Memo:
K
KAMEZAWA Hiroyuki 已提交
521 522
	recent_rotated means recent frequency of LRU rotation.
	recent_scanned means recent # of scans to LRU.
K
KOSAKI Motohiro 已提交
523 524
	showing for better debug please see the code for meanings.

B
Bharata B Rao 已提交
525 526 527
Note:
	Only anonymous and swap cache memory is listed as part of 'rss' stat.
	This should not be confused with the true 'resident set size' or the
K
KAMEZAWA Hiroyuki 已提交
528 529 530 531 532
	amount of physical memory used by the cgroup.
	'rss + file_mapped" will give you resident set size of cgroup.
	(Note: file and shmem may be shared among other cgroups. In that case,
	 file_mapped is accounted only when the memory cgroup is owner of page
	 cache.)
K
KOSAKI Motohiro 已提交
533

K
KOSAKI Motohiro 已提交
534 535
5.3 swappiness

K
KAMEZAWA Hiroyuki 已提交
536
Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only.
537 538 539 540
Please note that unlike the global swappiness, memcg knob set to 0
really prevents from any swapping even if there is a swap storage
available. This might lead to memcg OOM killer if there are no file
pages to reclaim.
K
KOSAKI Motohiro 已提交
541

K
KAMEZAWA Hiroyuki 已提交
542 543 544 545 546 547 548 549 550 551 552 553 554 555
Following cgroups' swappiness can't be changed.
- root cgroup (uses /proc/sys/vm/swappiness).
- a cgroup which uses hierarchy and it has other cgroup(s) below it.
- a cgroup which uses hierarchy and not the root of hierarchy.

5.4 failcnt

A memory cgroup provides memory.failcnt and memory.memsw.failcnt files.
This failcnt(== failure count) shows the number of times that a usage counter
hit its limit. When a memory cgroup hits a limit, failcnt increases and
memory under it will be reclaimed.

You can reset failcnt by writing 0 to failcnt file.
# echo 0 > .../memory.failcnt
K
KOSAKI Motohiro 已提交
556

557 558 559 560
5.5 usage_in_bytes

For efficiency, as other kernel components, memory cgroup uses some optimization
to avoid unnecessary cacheline false sharing. usage_in_bytes is affected by the
561
method and doesn't show 'exact' value of memory (and swap) usage, it's a fuzz
562 563 564 565
value for efficient access. (Of course, when necessary, it's synchronized.)
If you want to know more exact memory usage, you should use RSS+CACHE(+SWAP)
value in memory.stat(see 5.2).

566 567 568 569 570
5.6 numa_stat

This is similar to numa_maps but operates on a per-memcg basis.  This is
useful for providing visibility into the numa locality information within
an memcg since the pages are allowed to be allocated from any physical
571 572
node.  One of the use cases is evaluating application performance by
combining this information with the application's CPU allocation.
573 574 575 576 577 578 579 580 581 582 583

We export "total", "file", "anon" and "unevictable" pages per-node for
each memcg.  The ouput format of memory.numa_stat is:

total=<total pages> N0=<node 0 pages> N1=<node 1 pages> ...
file=<total file pages> N0=<node 0 pages> N1=<node 1 pages> ...
anon=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ...
unevictable=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ...

And we have total = file + anon + unevictable.

584
6. Hierarchy support
585

586 587 588 589 590
The memory controller supports a deep hierarchy and hierarchical accounting.
The hierarchy is created by creating the appropriate cgroups in the
cgroup filesystem. Consider for example, the following cgroup filesystem
hierarchy

591
	       root
592
	     /  |   \
593 594 595 596 597
            /	|    \
	   a	b     c
		      | \
		      |  \
		      d   e
598 599 600

In the diagram above, with hierarchical accounting enabled, all memory
usage of e, is accounted to its ancestors up until the root (i.e, c and root),
K
KAMEZAWA Hiroyuki 已提交
601
that has memory.use_hierarchy enabled. If one of the ancestors goes over its
602 603 604 605 606
limit, the reclaim algorithm reclaims from the tasks in the ancestor and the
children of the ancestor.

6.1 Enabling hierarchical accounting and reclaim

K
KAMEZAWA Hiroyuki 已提交
607
A memory cgroup by default disables the hierarchy feature. Support
608 609 610 611 612 613 614 615
can be enabled by writing 1 to memory.use_hierarchy file of the root cgroup

# echo 1 > memory.use_hierarchy

The feature can be disabled by

# echo 0 > memory.use_hierarchy

616 617 618
NOTE1: Enabling/disabling will fail if either the cgroup already has other
       cgroups created below it, or if the parent cgroup has use_hierarchy
       enabled.
619

620
NOTE2: When panic_on_oom is set to "2", the whole system will panic in
K
KAMEZAWA Hiroyuki 已提交
621
       case of an OOM event in any cgroup.
622

623 624 625 626 627 628 629 630
7. Soft limits

Soft limits allow for greater sharing of memory. The idea behind soft limits
is to allow control groups to use as much of the memory as needed, provided

a. There is no memory contention
b. They do not exceed their hard limit

K
KAMEZAWA Hiroyuki 已提交
631
When the system detects memory contention or low memory, control groups
632 633 634 635
are pushed back to their soft limits. If the soft limit of each control
group is very high, they are pushed back as much as possible to make
sure that one control group does not starve the others of memory.

636
Please note that soft limits is a best-effort feature; it comes with
637 638
no guarantees, but it does its best to make sure that when memory is
heavily contended for, memory is allocated based on the soft limit
639
hints/setup. Currently soft limit based reclaim is set up such that
640 641 642 643 644
it gets invoked from balance_pgdat (kswapd).

7.1 Interface

Soft limits can be setup by using the following commands (in this example we
K
KAMEZAWA Hiroyuki 已提交
645
assume a soft limit of 256 MiB)
646 647 648 649 650 651 652 653 654 655 656 657

# echo 256M > memory.soft_limit_in_bytes

If we want to change this to 1G, we can at any time use

# echo 1G > memory.soft_limit_in_bytes

NOTE1: Soft limits take effect over a long period of time, since they involve
       reclaiming memory for balancing between memory cgroups
NOTE2: It is recommended to set the soft limit always below the hard limit,
       otherwise the hard limit will take precedence.

658 659 660 661
8. Move charges at task migration

Users can move charges associated with a task along with task migration, that
is, uncharge task's pages from the old cgroup and charge them to the new cgroup.
662 663
This feature is not supported in !CONFIG_MMU environments because of lack of
page tables.
664 665 666

8.1 Interface

667
This feature is disabled by default. It can be enabledi (and disabled again) by
668 669 670 671 672 673 674 675
writing to memory.move_charge_at_immigrate of the destination cgroup.

If you want to enable it:

# echo (some positive value) > memory.move_charge_at_immigrate

Note: Each bits of move_charge_at_immigrate has its own meaning about what type
      of charges should be moved. See 8.2 for details.
676 677
Note: Charges are moved only when you move mm->owner, in other words,
      a leader of a thread group.
678 679 680
Note: If we cannot find enough space for the task in the destination cgroup, we
      try to make space by reclaiming memory. Task migration may fail if we
      cannot make enough space.
K
KAMEZAWA Hiroyuki 已提交
681
Note: It can take several seconds if you move charges much.
682 683 684 685 686

And if you want disable it again:

# echo 0 > memory.move_charge_at_immigrate

687
8.2 Type of charges which can be moved
688

689 690 691 692
Each bit in move_charge_at_immigrate has its own meaning about what type of
charges should be moved. But in any case, it must be noted that an account of
a page or a swap can be moved only when it is charged to the task's current
(old) memory cgroup.
693 694 695

  bit | what type of charges would be moved ?
 -----+------------------------------------------------------------------------
696 697
   0  | A charge of an anonymous page (or swap of it) used by the target task.
      | You must enable Swap Extension (see 2.4) to enable move of swap charges.
698
 -----+------------------------------------------------------------------------
699
   1  | A charge of file pages (normal file, tmpfs file (e.g. ipc shared memory)
K
KAMEZAWA Hiroyuki 已提交
700
      | and swaps of tmpfs file) mmapped by the target task. Unlike the case of
701
      | anonymous pages, file pages (and swaps) in the range mmapped by the task
702 703
      | will be moved even if the task hasn't done page fault, i.e. they might
      | not be the task's "RSS", but other task's "RSS" that maps the same file.
704 705
      | And mapcount of the page is ignored (the page can be moved even if
      | page_mapcount(page) > 1). You must enable Swap Extension (see 2.4) to
706
      | enable move of swap charges.
707 708 709 710 711 712

8.3 TODO

- All of moving charge operations are done under cgroup_mutex. It's not good
  behavior to hold the mutex too long, so we may need some trick.

713 714
9. Memory thresholds

715
Memory cgroup implements memory thresholds using the cgroups notification
716 717 718
API (see cgroups.txt). It allows to register multiple memory and memsw
thresholds and gets notifications when it crosses.

719
To register a threshold, an application must:
K
KAMEZAWA Hiroyuki 已提交
720 721 722 723
- create an eventfd using eventfd(2);
- open memory.usage_in_bytes or memory.memsw.usage_in_bytes;
- write string like "<event_fd> <fd of memory.usage_in_bytes> <threshold>" to
  cgroup.event_control.
724 725 726 727 728 729

Application will be notified through eventfd when memory usage crosses
threshold in any direction.

It's applicable for root and non-root cgroup.

K
KAMEZAWA Hiroyuki 已提交
730 731
10. OOM Control

732 733
memory.oom_control file is for OOM notification and other controls.

734
Memory cgroup implements OOM notifier using the cgroup notification
K
KAMEZAWA Hiroyuki 已提交
735 736
API (See cgroups.txt). It allows to register multiple OOM notification
delivery and gets notification when OOM happens.
K
KAMEZAWA Hiroyuki 已提交
737

738
To register a notifier, an application must:
K
KAMEZAWA Hiroyuki 已提交
739 740
 - create an eventfd using eventfd(2)
 - open memory.oom_control file
K
KAMEZAWA Hiroyuki 已提交
741 742
 - write string like "<event_fd> <fd of memory.oom_control>" to
   cgroup.event_control
K
KAMEZAWA Hiroyuki 已提交
743

744 745
The application will be notified through eventfd when OOM happens.
OOM notification doesn't work for the root cgroup.
K
KAMEZAWA Hiroyuki 已提交
746

747
You can disable the OOM-killer by writing "1" to memory.oom_control file, as:
K
KAMEZAWA Hiroyuki 已提交
748

749 750
	#echo 1 > memory.oom_control

751
This operation is only allowed to the top cgroup of a sub-hierarchy.
K
KAMEZAWA Hiroyuki 已提交
752 753
If OOM-killer is disabled, tasks under cgroup will hang/sleep
in memory cgroup's OOM-waitqueue when they request accountable memory.
754

K
KAMEZAWA Hiroyuki 已提交
755
For running them, you have to relax the memory cgroup's OOM status by
756 757 758 759 760 761 762 763 764 765
	* enlarge limit or reduce usage.
To reduce usage,
	* kill some tasks.
	* move some tasks to other group with account migration.
	* remove some files (on tmpfs?)

Then, stopped tasks will work again.

At reading, current status of OOM is shown.
	oom_kill_disable 0 or 1 (if 1, oom-killer is disabled)
K
KAMEZAWA Hiroyuki 已提交
766
	under_oom	 0 or 1 (if 1, the memory cgroup is under OOM, tasks may
767
				 be stopped.)
K
KAMEZAWA Hiroyuki 已提交
768

769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835
11. Memory Pressure

The pressure level notifications can be used to monitor the memory
allocation cost; based on the pressure, applications can implement
different strategies of managing their memory resources. The pressure
levels are defined as following:

The "low" level means that the system is reclaiming memory for new
allocations. Monitoring this reclaiming activity might be useful for
maintaining cache level. Upon notification, the program (typically
"Activity Manager") might analyze vmstat and act in advance (i.e.
prematurely shutdown unimportant services).

The "medium" level means that the system is experiencing medium memory
pressure, the system might be making swap, paging out active file caches,
etc. Upon this event applications may decide to further analyze
vmstat/zoneinfo/memcg or internal memory usage statistics and free any
resources that can be easily reconstructed or re-read from a disk.

The "critical" level means that the system is actively thrashing, it is
about to out of memory (OOM) or even the in-kernel OOM killer is on its
way to trigger. Applications should do whatever they can to help the
system. It might be too late to consult with vmstat or any other
statistics, so it's advisable to take an immediate action.

The events are propagated upward until the event is handled, i.e. the
events are not pass-through. Here is what this means: for example you have
three cgroups: A->B->C. Now you set up an event listener on cgroups A, B
and C, and suppose group C experiences some pressure. In this situation,
only group C will receive the notification, i.e. groups A and B will not
receive it. This is done to avoid excessive "broadcasting" of messages,
which disturbs the system and which is especially bad if we are low on
memory or thrashing. So, organize the cgroups wisely, or propagate the
events manually (or, ask us to implement the pass-through events,
explaining why would you need them.)

The file memory.pressure_level is only used to setup an eventfd. To
register a notification, an application must:

- create an eventfd using eventfd(2);
- open memory.pressure_level;
- write string like "<event_fd> <fd of memory.pressure_level> <level>"
  to cgroup.event_control.

Application will be notified through eventfd when memory pressure is at
the specific level (or higher). Read/write operations to
memory.pressure_level are no implemented.

Test:

   Here is a small script example that makes a new cgroup, sets up a
   memory limit, sets up a notification in the cgroup and then makes child
   cgroup experience a critical pressure:

   # cd /sys/fs/cgroup/memory/
   # mkdir foo
   # cd foo
   # cgroup_event_listener memory.pressure_level low &
   # echo 8000000 > memory.limit_in_bytes
   # echo 8000000 > memory.memsw.limit_in_bytes
   # echo $$ > tasks
   # dd if=/dev/zero | read x

   (Expect a bunch of notifications, and eventually, the oom-killer will
   trigger.)

12. TODO
B
Balbir Singh 已提交
836

837 838 839
1. Make per-cgroup scanner reclaim not-shared pages first
2. Teach controller to account for shared-pages
3. Start reclamation in the background when the limit is
B
Balbir Singh 已提交
840 841 842 843 844 845 846 847 848 849 850 851 852 853 854
   not yet hit but the usage is getting closer

Summary

Overall, the memory controller has been a stable controller and has been
commented and discussed quite extensively in the community.

References

1. Singh, Balbir. RFC: Memory Controller, http://lwn.net/Articles/206697/
2. Singh, Balbir. Memory Controller (RSS Control),
   http://lwn.net/Articles/222762/
3. Emelianov, Pavel. Resource controllers based on process cgroups
   http://lkml.org/lkml/2007/3/6/198
4. Emelianov, Pavel. RSS controller based on process cgroups (v2)
855
   http://lkml.org/lkml/2007/4/9/78
B
Balbir Singh 已提交
856 857 858 859 860
5. Emelianov, Pavel. RSS controller based on process cgroups (v3)
   http://lkml.org/lkml/2007/5/30/244
6. Menage, Paul. Control Groups v10, http://lwn.net/Articles/236032/
7. Vaidyanathan, Srinivasan, Control Groups: Pagecache accounting and control
   subsystem (v3), http://lwn.net/Articles/235534/
861
8. Singh, Balbir. RSS controller v2 test results (lmbench),
B
Balbir Singh 已提交
862
   http://lkml.org/lkml/2007/5/17/232
863
9. Singh, Balbir. RSS controller v2 AIM9 results
B
Balbir Singh 已提交
864
   http://lkml.org/lkml/2007/5/18/1
865
10. Singh, Balbir. Memory controller v6 test results,
B
Balbir Singh 已提交
866
    http://lkml.org/lkml/2007/8/19/36
867 868
11. Singh, Balbir. Memory controller introduction (v6),
    http://lkml.org/lkml/2007/8/17/69
B
Balbir Singh 已提交
869 870
12. Corbet, Jonathan, Controlling memory use in cgroups,
    http://lwn.net/Articles/243795/