]>
Commit | Line | Data |
---|---|---|
00f0b825 BS |
1 | Memory Resource Controller |
2 | ||
67de0162 JS |
3 | NOTE: The Memory Resource Controller has generically been referred to as the |
4 | memory controller in this document. Do not confuse memory controller | |
5 | used here with the memory controller that is used in hardware. | |
1b6df3aa | 6 | |
dc10e281 KH |
7 | (For editors) |
8 | In this document: | |
9 | When we mention a cgroup (cgroupfs's directory) with memory controller, | |
10 | we call it "memory cgroup". When you see git-log and source code, you'll | |
11 | see patch's title and function names tend to use "memcg". | |
12 | In this document, we avoid using it. | |
1b6df3aa | 13 | |
1b6df3aa BS |
14 | Benefits and Purpose of the memory controller |
15 | ||
16 | The memory controller isolates the memory behaviour of a group of tasks | |
17 | from the rest of the system. The article on LWN [12] mentions some probable | |
18 | uses of the memory controller. The memory controller can be used to | |
19 | ||
20 | a. Isolate an application or a group of applications | |
1939c557 | 21 | Memory-hungry applications can be isolated and limited to a smaller |
1b6df3aa | 22 | amount of memory. |
1939c557 | 23 | b. Create a cgroup with a limited amount of memory; this can be used |
1b6df3aa BS |
24 | as a good alternative to booting with mem=XXXX. |
25 | c. Virtualization solutions can control the amount of memory they want | |
26 | to assign to a virtual machine instance. | |
27 | d. A CD/DVD burner could control the amount of memory used by the | |
28 | rest of the system to ensure that burning does not fail due to lack | |
29 | of available memory. | |
1939c557 | 30 | e. There are several other use cases; find one or use the controller just |
1b6df3aa BS |
31 | for fun (to learn and hack on the VM subsystem). |
32 | ||
dc10e281 KH |
33 | Current Status: linux-2.6.34-mmotm(development version of 2010/April) |
34 | ||
35 | Features: | |
36 | - accounting anonymous pages, file caches, swap caches usage and limiting them. | |
6252efcc | 37 | - pages are linked to per-memcg LRU exclusively, and there is no global LRU. |
dc10e281 KH |
38 | - optionally, memory+swap usage can be accounted and limited. |
39 | - hierarchical accounting | |
40 | - soft limit | |
1939c557 | 41 | - moving (recharging) account at moving a task is selectable. |
dc10e281 | 42 | - usage threshold notifier |
70ddf637 | 43 | - memory pressure notifier |
dc10e281 KH |
44 | - oom-killer disable knob and oom-notifier |
45 | - Root cgroup has no limit controls. | |
46 | ||
1939c557 | 47 | Kernel memory support is a work in progress, and the current version provides |
65c64ce8 | 48 | basically functionality. (See Section 2.7) |
dc10e281 KH |
49 | |
50 | Brief summary of control files. | |
51 | ||
52 | tasks # attach a task(thread) and show list of threads | |
53 | cgroup.procs # show list of processes | |
54 | cgroup.event_control # an interface for event_fd() | |
a111c966 DN |
55 | memory.usage_in_bytes # show current res_counter usage for memory |
56 | (See 5.5 for details) | |
57 | memory.memsw.usage_in_bytes # show current res_counter usage for memory+Swap | |
58 | (See 5.5 for details) | |
dc10e281 KH |
59 | memory.limit_in_bytes # set/show limit of memory usage |
60 | memory.memsw.limit_in_bytes # set/show limit of memory+Swap usage | |
61 | memory.failcnt # show the number of memory usage hits limits | |
62 | memory.memsw.failcnt # show the number of memory+Swap hits limits | |
63 | memory.max_usage_in_bytes # show max memory usage recorded | |
d66c1ce7 | 64 | memory.memsw.max_usage_in_bytes # show max memory+Swap usage recorded |
dc10e281 KH |
65 | memory.soft_limit_in_bytes # set/show soft limit of memory usage |
66 | memory.stat # show various statistics | |
67 | memory.use_hierarchy # set/show hierarchical account enabled | |
68 | memory.force_empty # trigger forced move charge to parent | |
70ddf637 | 69 | memory.pressure_level # set memory pressure notifications |
dc10e281 KH |
70 | memory.swappiness # set/show swappiness parameter of vmscan |
71 | (See sysctl's vm.swappiness) | |
72 | memory.move_charge_at_immigrate # set/show controls of moving charges | |
73 | memory.oom_control # set/show oom controls. | |
50c35e5b | 74 | memory.numa_stat # show the number of memory usage per numa node |
dc10e281 | 75 | |
d5bdae7d GC |
76 | memory.kmem.limit_in_bytes # set/show hard limit for kernel memory |
77 | memory.kmem.usage_in_bytes # show current kernel memory allocation | |
78 | memory.kmem.failcnt # show the number of kernel memory usage hits limits | |
79 | memory.kmem.max_usage_in_bytes # show max kernel memory usage recorded | |
80 | ||
3aaabe23 | 81 | memory.kmem.tcp.limit_in_bytes # set/show hard limit for tcp buf memory |
5a6dd343 | 82 | memory.kmem.tcp.usage_in_bytes # show current tcp buf memory allocation |
05a73ed2 WL |
83 | memory.kmem.tcp.failcnt # show the number of tcp buf memory usage hits limits |
84 | memory.kmem.tcp.max_usage_in_bytes # show max tcp buf memory usage recorded | |
e5671dfa | 85 | |
1b6df3aa BS |
86 | 1. History |
87 | ||
88 | The memory controller has a long history. A request for comments for the memory | |
89 | controller was posted by Balbir Singh [1]. At the time the RFC was posted | |
90 | there were several implementations for memory control. The goal of the | |
91 | RFC was to build consensus and agreement for the minimal features required | |
92 | for memory control. The first RSS controller was posted by Balbir Singh[2] | |
93 | in Feb 2007. Pavel Emelianov [3][4][5] has since posted three versions of the | |
94 | RSS controller. At OLS, at the resource management BoF, everyone suggested | |
95 | that we handle both page cache and RSS together. Another request was raised | |
96 | to allow user space handling of OOM. The current memory controller is | |
97 | at version 6; it combines both mapped (RSS) and unmapped Page | |
98 | Cache Control [11]. | |
99 | ||
100 | 2. Memory Control | |
101 | ||
102 | Memory is a unique resource in the sense that it is present in a limited | |
103 | amount. If a task requires a lot of CPU processing, the task can spread | |
104 | its processing over a period of hours, days, months or years, but with | |
105 | memory, the same physical memory needs to be reused to accomplish the task. | |
106 | ||
107 | The memory controller implementation has been divided into phases. These | |
108 | are: | |
109 | ||
110 | 1. Memory controller | |
111 | 2. mlock(2) controller | |
112 | 3. Kernel user memory accounting and slab control | |
113 | 4. user mappings length controller | |
114 | ||
115 | The memory controller is the first controller developed. | |
116 | ||
117 | 2.1. Design | |
118 | ||
119 | The core of the design is a counter called the res_counter. The res_counter | |
120 | tracks the current memory usage and limit of the group of processes associated | |
121 | with the controller. Each cgroup has a memory controller specific data | |
122 | structure (mem_cgroup) associated with it. | |
123 | ||
124 | 2.2. Accounting | |
125 | ||
126 | +--------------------+ | |
127 | | mem_cgroup | | |
128 | | (res_counter) | | |
129 | +--------------------+ | |
130 | / ^ \ | |
131 | / | \ | |
132 | +---------------+ | +---------------+ | |
133 | | mm_struct | |.... | mm_struct | | |
134 | | | | | | | |
135 | +---------------+ | +---------------+ | |
136 | | | |
137 | + --------------+ | |
138 | | | |
139 | +---------------+ +------+--------+ | |
140 | | page +----------> page_cgroup| | |
141 | | | | | | |
142 | +---------------+ +---------------+ | |
143 | ||
144 | (Figure 1: Hierarchy of Accounting) | |
145 | ||
146 | ||
147 | Figure 1 shows the important aspects of the controller | |
148 | ||
149 | 1. Accounting happens per cgroup | |
150 | 2. Each mm_struct knows about which cgroup it belongs to | |
151 | 3. Each page has a pointer to the page_cgroup, which in turn knows the | |
152 | cgroup it belongs to | |
153 | ||
348b4655 JL |
154 | The accounting is done as follows: mem_cgroup_charge_common() is invoked to |
155 | set up the necessary data structures and check if the cgroup that is being | |
156 | charged is over its limit. If it is, then reclaim is invoked on the cgroup. | |
1b6df3aa BS |
157 | More details can be found in the reclaim section of this document. |
158 | If everything goes well, a page meta-data-structure called page_cgroup is | |
dc10e281 KH |
159 | updated. page_cgroup has its own LRU on cgroup. |
160 | (*) page_cgroup structure is allocated at boot/memory-hotplug time. | |
1b6df3aa BS |
161 | |
162 | 2.2.1 Accounting details | |
163 | ||
5b4e655e | 164 | All mapped anon pages (RSS) and cache pages (Page Cache) are accounted. |
6252efcc | 165 | Some pages which are never reclaimable and will not be on the LRU |
dc10e281 | 166 | are not accounted. We just account pages under usual VM management. |
5b4e655e KH |
167 | |
168 | RSS pages are accounted at page_fault unless they've already been accounted | |
169 | for earlier. A file page will be accounted for as Page Cache when it's | |
170 | inserted into inode (radix-tree). While it's mapped into the page tables of | |
171 | processes, duplicate accounting is carefully avoided. | |
172 | ||
1939c557 | 173 | An RSS page is unaccounted when it's fully unmapped. A PageCache page is |
dc10e281 KH |
174 | unaccounted when it's removed from radix-tree. Even if RSS pages are fully |
175 | unmapped (by kswapd), they may exist as SwapCache in the system until they | |
1939c557 | 176 | are really freed. Such SwapCaches are also accounted. |
dc10e281 KH |
177 | A swapped-in page is not accounted until it's mapped. |
178 | ||
1939c557 | 179 | Note: The kernel does swapin-readahead and reads multiple swaps at once. |
dc10e281 KH |
180 | This means swapped-in pages may contain pages for other tasks than a task |
181 | causing page fault. So, we avoid accounting at swap-in I/O. | |
5b4e655e KH |
182 | |
183 | At page migration, accounting information is kept. | |
184 | ||
dc10e281 KH |
185 | Note: we just account pages-on-LRU because our purpose is to control amount |
186 | of used pages; not-on-LRU pages tend to be out-of-control from VM view. | |
1b6df3aa BS |
187 | |
188 | 2.3 Shared Page Accounting | |
189 | ||
190 | Shared pages are accounted on the basis of the first touch approach. The | |
191 | cgroup that first touches a page is accounted for the page. The principle | |
192 | behind this approach is that a cgroup that aggressively uses a shared | |
193 | page will eventually get charged for it (once it is uncharged from | |
194 | the cgroup that brought it in -- this will happen on memory pressure). | |
195 | ||
4b91355e KH |
196 | But see section 8.2: when moving a task to another cgroup, its pages may |
197 | be recharged to the new cgroup, if move_charge_at_immigrate has been chosen. | |
198 | ||
df7c6b99 | 199 | Exception: If CONFIG_MEMCG_SWAP is not used. |
8c7c6e34 | 200 | When you do swapoff and make swapped-out pages of shmem(tmpfs) to |
d13d1443 KH |
201 | be backed into memory in force, charges for pages are accounted against the |
202 | caller of swapoff rather than the users of shmem. | |
203 | ||
c255a458 | 204 | 2.4 Swap Extension (CONFIG_MEMCG_SWAP) |
dc10e281 | 205 | |
8c7c6e34 KH |
206 | Swap Extension allows you to record charge for swap. A swapped-in page is |
207 | charged back to original page allocator if possible. | |
208 | ||
209 | When swap is accounted, following files are added. | |
210 | - memory.memsw.usage_in_bytes. | |
211 | - memory.memsw.limit_in_bytes. | |
212 | ||
dc10e281 KH |
213 | memsw means memory+swap. Usage of memory+swap is limited by |
214 | memsw.limit_in_bytes. | |
215 | ||
216 | Example: Assume a system with 4G of swap. A task which allocates 6G of memory | |
217 | (by mistake) under 2G memory limitation will use all swap. | |
218 | In this case, setting memsw.limit_in_bytes=3G will prevent bad use of swap. | |
1939c557 | 219 | By using the memsw limit, you can avoid system OOM which can be caused by swap |
dc10e281 | 220 | shortage. |
8c7c6e34 | 221 | |
dc10e281 | 222 | * why 'memory+swap' rather than swap. |
8c7c6e34 KH |
223 | The global LRU(kswapd) can swap out arbitrary pages. Swap-out means |
224 | to move account from memory to swap...there is no change in usage of | |
dc10e281 KH |
225 | memory+swap. In other words, when we want to limit the usage of swap without |
226 | affecting global LRU, memory+swap limit is better than just limiting swap from | |
1939c557 | 227 | an OS point of view. |
22a668d7 KH |
228 | |
229 | * What happens when a cgroup hits memory.memsw.limit_in_bytes | |
67de0162 | 230 | When a cgroup hits memory.memsw.limit_in_bytes, it's useless to do swap-out |
22a668d7 KH |
231 | in this cgroup. Then, swap-out will not be done by cgroup routine and file |
232 | caches are dropped. But as mentioned above, global LRU can do swapout memory | |
233 | from it for sanity of the system's memory management state. You can't forbid | |
234 | it by cgroup. | |
8c7c6e34 KH |
235 | |
236 | 2.5 Reclaim | |
1b6df3aa | 237 | |
dc10e281 KH |
238 | Each cgroup maintains a per cgroup LRU which has the same structure as |
239 | global VM. When a cgroup goes over its limit, we first try | |
1b6df3aa BS |
240 | to reclaim memory from the cgroup so as to make space for the new |
241 | pages that the cgroup has touched. If the reclaim is unsuccessful, | |
242 | an OOM routine is invoked to select and kill the bulkiest task in the | |
dc10e281 | 243 | cgroup. (See 10. OOM Control below.) |
1b6df3aa BS |
244 | |
245 | The reclaim algorithm has not been modified for cgroups, except that | |
1939c557 | 246 | pages that are selected for reclaiming come from the per-cgroup LRU |
1b6df3aa BS |
247 | list. |
248 | ||
4b3bde4c BS |
249 | NOTE: Reclaim does not work for the root cgroup, since we cannot set any |
250 | limits on the root cgroup. | |
251 | ||
daaf1e68 KH |
252 | Note2: When panic_on_oom is set to "2", the whole system will panic. |
253 | ||
9490ff27 KH |
254 | When oom event notifier is registered, event will be delivered. |
255 | (See oom_control section) | |
256 | ||
dc10e281 | 257 | 2.6 Locking |
1b6df3aa | 258 | |
dc10e281 KH |
259 | lock_page_cgroup()/unlock_page_cgroup() should not be called under |
260 | mapping->tree_lock. | |
1b6df3aa | 261 | |
dc10e281 KH |
262 | Other lock order is following: |
263 | PG_locked. | |
264 | mm->page_table_lock | |
265 | zone->lru_lock | |
266 | lock_page_cgroup. | |
267 | In many cases, just lock_page_cgroup() is called. | |
268 | per-zone-per-cgroup LRU (cgroup's private LRU) is just guarded by | |
269 | zone->lru_lock, it has no lock of its own. | |
1b6df3aa | 270 | |
c255a458 | 271 | 2.7 Kernel Memory Extension (CONFIG_MEMCG_KMEM) |
e5671dfa GC |
272 | |
273 | With the Kernel memory extension, the Memory Controller is able to limit | |
274 | the amount of kernel memory used by the system. Kernel memory is fundamentally | |
275 | different than user memory, since it can't be swapped out, which makes it | |
276 | possible to DoS the system by consuming too much of this precious resource. | |
277 | ||
d5bdae7d GC |
278 | Kernel memory won't be accounted at all until limit on a group is set. This |
279 | allows for existing setups to continue working without disruption. The limit | |
280 | cannot be set if the cgroup have children, or if there are already tasks in the | |
281 | cgroup. Attempting to set the limit under those conditions will return -EBUSY. | |
282 | When use_hierarchy == 1 and a group is accounted, its children will | |
283 | automatically be accounted regardless of their limit value. | |
284 | ||
285 | After a group is first limited, it will be kept being accounted until it | |
286 | is removed. The memory limitation itself, can of course be removed by writing | |
287 | -1 to memory.kmem.limit_in_bytes. In this case, kmem will be accounted, but not | |
288 | limited. | |
289 | ||
e5671dfa | 290 | Kernel memory limits are not imposed for the root cgroup. Usage for the root |
d5bdae7d GC |
291 | cgroup may or may not be accounted. The memory used is accumulated into |
292 | memory.kmem.usage_in_bytes, or in a separate counter when it makes sense. | |
293 | (currently only for tcp). | |
294 | The main "kmem" counter is fed into the main counter, so kmem charges will | |
295 | also be visible from the user counter. | |
e5671dfa | 296 | |
e5671dfa GC |
297 | Currently no soft limit is implemented for kernel memory. It is future work |
298 | to trigger slab reclaim when those limits are reached. | |
299 | ||
300 | 2.7.1 Current Kernel Memory resources accounted | |
301 | ||
d5bdae7d GC |
302 | * stack pages: every process consumes some stack pages. By accounting into |
303 | kernel memory, we prevent new processes from being created when the kernel | |
304 | memory usage is too high. | |
305 | ||
92e79349 GC |
306 | * slab pages: pages allocated by the SLAB or SLUB allocator are tracked. A copy |
307 | of each kmem_cache is created everytime the cache is touched by the first time | |
308 | from inside the memcg. The creation is done lazily, so some objects can still be | |
309 | skipped while the cache is being created. All objects in a slab page should | |
310 | belong to the same memcg. This only fails to hold when a task is migrated to a | |
311 | different memcg during the page allocation by the cache. | |
312 | ||
e1aab161 GC |
313 | * sockets memory pressure: some sockets protocols have memory pressure |
314 | thresholds. The Memory Controller allows them to be controlled individually | |
315 | per cgroup, instead of globally. | |
e5671dfa | 316 | |
d1a4c0b3 GC |
317 | * tcp memory pressure: sockets memory pressure for the tcp protocol. |
318 | ||
d5bdae7d GC |
319 | 2.7.3 Common use cases |
320 | ||
321 | Because the "kmem" counter is fed to the main user counter, kernel memory can | |
322 | never be limited completely independently of user memory. Say "U" is the user | |
323 | limit, and "K" the kernel limit. There are three possible ways limits can be | |
324 | set: | |
325 | ||
326 | U != 0, K = unlimited: | |
327 | This is the standard memcg limitation mechanism already present before kmem | |
328 | accounting. Kernel memory is completely ignored. | |
329 | ||
330 | U != 0, K < U: | |
331 | Kernel memory is a subset of the user memory. This setup is useful in | |
332 | deployments where the total amount of memory per-cgroup is overcommited. | |
333 | Overcommiting kernel memory limits is definitely not recommended, since the | |
334 | box can still run out of non-reclaimable memory. | |
335 | In this case, the admin could set up K so that the sum of all groups is | |
336 | never greater than the total memory, and freely set U at the cost of his | |
337 | QoS. | |
338 | ||
339 | U != 0, K >= U: | |
340 | Since kmem charges will also be fed to the user counter and reclaim will be | |
341 | triggered for the cgroup for both kinds of memory. This setup gives the | |
342 | admin a unified view of memory, and it is also useful for people who just | |
343 | want to track kernel memory usage. | |
344 | ||
1b6df3aa BS |
345 | 3. User Interface |
346 | ||
347 | 0. Configuration | |
348 | ||
349 | a. Enable CONFIG_CGROUPS | |
350 | b. Enable CONFIG_RESOURCE_COUNTERS | |
c255a458 AM |
351 | c. Enable CONFIG_MEMCG |
352 | d. Enable CONFIG_MEMCG_SWAP (to use swap extension) | |
d5bdae7d | 353 | d. Enable CONFIG_MEMCG_KMEM (to use kmem extension) |
1b6df3aa | 354 | |
f6e07d38 JS |
355 | 1. Prepare the cgroups (see cgroups.txt, Why are cgroups needed?) |
356 | # mount -t tmpfs none /sys/fs/cgroup | |
357 | # mkdir /sys/fs/cgroup/memory | |
358 | # mount -t cgroup none /sys/fs/cgroup/memory -o memory | |
1b6df3aa BS |
359 | |
360 | 2. Make the new group and move bash into it | |
f6e07d38 JS |
361 | # mkdir /sys/fs/cgroup/memory/0 |
362 | # echo $$ > /sys/fs/cgroup/memory/0/tasks | |
1b6df3aa | 363 | |
dc10e281 | 364 | Since now we're in the 0 cgroup, we can alter the memory limit: |
f6e07d38 | 365 | # echo 4M > /sys/fs/cgroup/memory/0/memory.limit_in_bytes |
0eea1030 BS |
366 | |
367 | NOTE: We can use a suffix (k, K, m, M, g or G) to indicate values in kilo, | |
dc10e281 KH |
368 | mega or gigabytes. (Here, Kilo, Mega, Giga are Kibibytes, Mebibytes, Gibibytes.) |
369 | ||
c5b947b2 | 370 | NOTE: We can write "-1" to reset the *.limit_in_bytes(unlimited). |
4b3bde4c | 371 | NOTE: We cannot set limits on the root cgroup any more. |
0eea1030 | 372 | |
f6e07d38 | 373 | # cat /sys/fs/cgroup/memory/0/memory.limit_in_bytes |
2324c5dd | 374 | 4194304 |
0eea1030 | 375 | |
1b6df3aa | 376 | We can check the usage: |
f6e07d38 | 377 | # cat /sys/fs/cgroup/memory/0/memory.usage_in_bytes |
2324c5dd | 378 | 1216512 |
0eea1030 | 379 | |
1939c557 | 380 | A successful write to this file does not guarantee a successful setting of |
dc10e281 | 381 | this limit to the value written into the file. This can be due to a |
0eea1030 | 382 | number of factors, such as rounding up to page boundaries or the total |
dc10e281 | 383 | availability of memory on the system. The user is required to re-read |
0eea1030 BS |
384 | this file after a write to guarantee the value committed by the kernel. |
385 | ||
fb78922c | 386 | # echo 1 > memory.limit_in_bytes |
0eea1030 | 387 | # cat memory.limit_in_bytes |
2324c5dd | 388 | 4096 |
1b6df3aa BS |
389 | |
390 | The memory.failcnt field gives the number of times that the cgroup limit was | |
391 | exceeded. | |
392 | ||
dfc05c25 KH |
393 | The memory.stat file gives accounting information. Now, the number of |
394 | caches, RSS and Active pages/Inactive pages are shown. | |
395 | ||
1b6df3aa BS |
396 | 4. Testing |
397 | ||
dc10e281 KH |
398 | For testing features and implementation, see memcg_test.txt. |
399 | ||
400 | Performance test is also important. To see pure memory controller's overhead, | |
401 | testing on tmpfs will give you good numbers of small overheads. | |
402 | Example: do kernel make on tmpfs. | |
403 | ||
404 | Page-fault scalability is also important. At measuring parallel | |
405 | page fault test, multi-process test may be better than multi-thread | |
406 | test because it has noise of shared objects/status. | |
407 | ||
408 | But the above two are testing extreme situations. | |
409 | Trying usual test under memory controller is always helpful. | |
1b6df3aa BS |
410 | |
411 | 4.1 Troubleshooting | |
412 | ||
413 | Sometimes a user might find that the application under a cgroup is | |
1939c557 | 414 | terminated by the OOM killer. There are several causes for this: |
1b6df3aa BS |
415 | |
416 | 1. The cgroup limit is too low (just too low to do anything useful) | |
417 | 2. The user is using anonymous memory and swap is turned off or too low | |
418 | ||
419 | A sync followed by echo 1 > /proc/sys/vm/drop_caches will help get rid of | |
420 | some of the pages cached in the cgroup (page cache pages). | |
421 | ||
1939c557 | 422 | To know what happens, disabling OOM_Kill as per "10. OOM Control" (below) and |
dc10e281 KH |
423 | seeing what happens will be helpful. |
424 | ||
1b6df3aa BS |
425 | 4.2 Task migration |
426 | ||
a33f3224 | 427 | When a task migrates from one cgroup to another, its charge is not |
7dc74be0 | 428 | carried forward by default. The pages allocated from the original cgroup still |
1b6df3aa BS |
429 | remain charged to it, the charge is dropped when the page is freed or |
430 | reclaimed. | |
431 | ||
dc10e281 KH |
432 | You can move charges of a task along with task migration. |
433 | See 8. "Move charges at task migration" | |
7dc74be0 | 434 | |
1b6df3aa BS |
435 | 4.3 Removing a cgroup |
436 | ||
437 | A cgroup can be removed by rmdir, but as discussed in sections 4.1 and 4.2, a | |
438 | cgroup might have some charge associated with it, even though all | |
dc10e281 KH |
439 | tasks have migrated away from it. (because we charge against pages, not |
440 | against tasks.) | |
441 | ||
cc926f78 KH |
442 | We move the stats to root (if use_hierarchy==0) or parent (if |
443 | use_hierarchy==1), and no change on the charge except uncharging | |
444 | from the child. | |
1b6df3aa | 445 | |
8c7c6e34 KH |
446 | Charges recorded in swap information is not updated at removal of cgroup. |
447 | Recorded information is discarded and a cgroup which uses swap (swapcache) | |
448 | will be charged as a new owner of it. | |
449 | ||
cc926f78 | 450 | About use_hierarchy, see Section 6. |
8c7c6e34 | 451 | |
c1e862c1 KH |
452 | 5. Misc. interfaces. |
453 | ||
454 | 5.1 force_empty | |
455 | memory.force_empty interface is provided to make cgroup's memory usage empty. | |
456 | You can use this interface only when the cgroup has no tasks. | |
457 | When writing anything to this | |
458 | ||
459 | # echo 0 > memory.force_empty | |
460 | ||
dc10e281 KH |
461 | Almost all pages tracked by this memory cgroup will be unmapped and freed. |
462 | Some pages cannot be freed because they are locked or in-use. Such pages are | |
1939c557 | 463 | moved to parent (if use_hierarchy==1) or root (if use_hierarchy==0) and this |
cc926f78 | 464 | cgroup will be empty. |
c1e862c1 | 465 | |
1939c557 | 466 | The typical use case for this interface is before calling rmdir(). |
c1e862c1 KH |
467 | Because rmdir() moves all pages to parent, some out-of-use page caches can be |
468 | moved to the parent. If you want to avoid that, force_empty will be useful. | |
469 | ||
d5bdae7d GC |
470 | Also, note that when memory.kmem.limit_in_bytes is set the charges due to |
471 | kernel pages will still be seen. This is not considered a failure and the | |
472 | write will still return success. In this case, it is expected that | |
473 | memory.kmem.usage_in_bytes == memory.usage_in_bytes. | |
474 | ||
cc926f78 KH |
475 | About use_hierarchy, see Section 6. |
476 | ||
7f016ee8 | 477 | 5.2 stat file |
c863d835 | 478 | |
185efc0f | 479 | memory.stat file includes following statistics |
c863d835 | 480 | |
dc10e281 | 481 | # per-memory cgroup local status |
c863d835 | 482 | cache - # of bytes of page cache memory. |
b070e65c DR |
483 | rss - # of bytes of anonymous and swap cache memory (includes |
484 | transparent hugepages). | |
485 | rss_huge - # of bytes of anonymous transparent hugepages. | |
dc10e281 | 486 | mapped_file - # of bytes of mapped file (includes tmpfs/shmem) |
0527b690 YH |
487 | pgpgin - # of charging events to the memory cgroup. The charging |
488 | event happens each time a page is accounted as either mapped | |
489 | anon page(RSS) or cache page(Page Cache) to the cgroup. | |
490 | pgpgout - # of uncharging events to the memory cgroup. The uncharging | |
491 | event happens each time a page is unaccounted from the cgroup. | |
dc10e281 | 492 | swap - # of bytes of swap usage |
c863d835 | 493 | inactive_anon - # of bytes of anonymous memory and swap cache memory on |
dc10e281 KH |
494 | LRU list. |
495 | active_anon - # of bytes of anonymous and swap cache memory on active | |
496 | inactive LRU list. | |
497 | inactive_file - # of bytes of file-backed memory on inactive LRU list. | |
498 | active_file - # of bytes of file-backed memory on active LRU list. | |
c863d835 BR |
499 | unevictable - # of bytes of memory that cannot be reclaimed (mlocked etc). |
500 | ||
dc10e281 KH |
501 | # status considering hierarchy (see memory.use_hierarchy settings) |
502 | ||
503 | hierarchical_memory_limit - # of bytes of memory limit with regard to hierarchy | |
504 | under which the memory cgroup is | |
505 | hierarchical_memsw_limit - # of bytes of memory+swap limit with regard to | |
506 | hierarchy under which memory cgroup is. | |
507 | ||
eb6332a5 JW |
508 | total_<counter> - # hierarchical version of <counter>, which in |
509 | addition to the cgroup's own value includes the | |
510 | sum of all hierarchical children's values of | |
511 | <counter>, i.e. total_cache | |
dc10e281 KH |
512 | |
513 | # The following additional stats are dependent on CONFIG_DEBUG_VM. | |
c863d835 | 514 | |
c863d835 BR |
515 | recent_rotated_anon - VM internal parameter. (see mm/vmscan.c) |
516 | recent_rotated_file - VM internal parameter. (see mm/vmscan.c) | |
517 | recent_scanned_anon - VM internal parameter. (see mm/vmscan.c) | |
518 | recent_scanned_file - VM internal parameter. (see mm/vmscan.c) | |
519 | ||
520 | Memo: | |
dc10e281 KH |
521 | recent_rotated means recent frequency of LRU rotation. |
522 | recent_scanned means recent # of scans to LRU. | |
7f016ee8 KM |
523 | showing for better debug please see the code for meanings. |
524 | ||
c863d835 BR |
525 | Note: |
526 | Only anonymous and swap cache memory is listed as part of 'rss' stat. | |
527 | This should not be confused with the true 'resident set size' or the | |
dc10e281 KH |
528 | amount of physical memory used by the cgroup. |
529 | 'rss + file_mapped" will give you resident set size of cgroup. | |
530 | (Note: file and shmem may be shared among other cgroups. In that case, | |
531 | file_mapped is accounted only when the memory cgroup is owner of page | |
532 | cache.) | |
7f016ee8 | 533 | |
a7885eb8 | 534 | 5.3 swappiness |
a7885eb8 | 535 | |
dc10e281 | 536 | Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only. |
9a5a8f19 MH |
537 | Please note that unlike the global swappiness, memcg knob set to 0 |
538 | really prevents from any swapping even if there is a swap storage | |
539 | available. This might lead to memcg OOM killer if there are no file | |
540 | pages to reclaim. | |
a7885eb8 | 541 | |
dc10e281 KH |
542 | Following cgroups' swappiness can't be changed. |
543 | - root cgroup (uses /proc/sys/vm/swappiness). | |
544 | - a cgroup which uses hierarchy and it has other cgroup(s) below it. | |
545 | - a cgroup which uses hierarchy and not the root of hierarchy. | |
546 | ||
547 | 5.4 failcnt | |
548 | ||
549 | A memory cgroup provides memory.failcnt and memory.memsw.failcnt files. | |
550 | This failcnt(== failure count) shows the number of times that a usage counter | |
551 | hit its limit. When a memory cgroup hits a limit, failcnt increases and | |
552 | memory under it will be reclaimed. | |
553 | ||
554 | You can reset failcnt by writing 0 to failcnt file. | |
555 | # echo 0 > .../memory.failcnt | |
a7885eb8 | 556 | |
a111c966 DN |
557 | 5.5 usage_in_bytes |
558 | ||
559 | For efficiency, as other kernel components, memory cgroup uses some optimization | |
560 | to avoid unnecessary cacheline false sharing. usage_in_bytes is affected by the | |
1939c557 | 561 | method and doesn't show 'exact' value of memory (and swap) usage, it's a fuzz |
a111c966 DN |
562 | value for efficient access. (Of course, when necessary, it's synchronized.) |
563 | If you want to know more exact memory usage, you should use RSS+CACHE(+SWAP) | |
564 | value in memory.stat(see 5.2). | |
565 | ||
50c35e5b YH |
566 | 5.6 numa_stat |
567 | ||
568 | This is similar to numa_maps but operates on a per-memcg basis. This is | |
569 | useful for providing visibility into the numa locality information within | |
570 | an memcg since the pages are allowed to be allocated from any physical | |
1939c557 MK |
571 | node. One of the use cases is evaluating application performance by |
572 | combining this information with the application's CPU allocation. | |
50c35e5b YH |
573 | |
574 | We export "total", "file", "anon" and "unevictable" pages per-node for | |
575 | each memcg. The ouput format of memory.numa_stat is: | |
576 | ||
577 | total=<total pages> N0=<node 0 pages> N1=<node 1 pages> ... | |
578 | file=<total file pages> N0=<node 0 pages> N1=<node 1 pages> ... | |
579 | anon=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ... | |
580 | unevictable=<total anon pages> N0=<node 0 pages> N1=<node 1 pages> ... | |
581 | ||
582 | And we have total = file + anon + unevictable. | |
583 | ||
52bc0d82 | 584 | 6. Hierarchy support |
c1e862c1 | 585 | |
52bc0d82 BS |
586 | The memory controller supports a deep hierarchy and hierarchical accounting. |
587 | The hierarchy is created by creating the appropriate cgroups in the | |
588 | cgroup filesystem. Consider for example, the following cgroup filesystem | |
589 | hierarchy | |
590 | ||
67de0162 | 591 | root |
52bc0d82 | 592 | / | \ |
67de0162 JS |
593 | / | \ |
594 | a b c | |
595 | | \ | |
596 | | \ | |
597 | d e | |
52bc0d82 BS |
598 | |
599 | In the diagram above, with hierarchical accounting enabled, all memory | |
600 | usage of e, is accounted to its ancestors up until the root (i.e, c and root), | |
dc10e281 | 601 | that has memory.use_hierarchy enabled. If one of the ancestors goes over its |
52bc0d82 BS |
602 | limit, the reclaim algorithm reclaims from the tasks in the ancestor and the |
603 | children of the ancestor. | |
604 | ||
605 | 6.1 Enabling hierarchical accounting and reclaim | |
606 | ||
dc10e281 | 607 | A memory cgroup by default disables the hierarchy feature. Support |
52bc0d82 BS |
608 | can be enabled by writing 1 to memory.use_hierarchy file of the root cgroup |
609 | ||
610 | # echo 1 > memory.use_hierarchy | |
611 | ||
612 | The feature can be disabled by | |
613 | ||
614 | # echo 0 > memory.use_hierarchy | |
615 | ||
689bca3b GT |
616 | NOTE1: Enabling/disabling will fail if either the cgroup already has other |
617 | cgroups created below it, or if the parent cgroup has use_hierarchy | |
618 | enabled. | |
52bc0d82 | 619 | |
daaf1e68 | 620 | NOTE2: When panic_on_oom is set to "2", the whole system will panic in |
dc10e281 | 621 | case of an OOM event in any cgroup. |
52bc0d82 | 622 | |
a6df6361 BS |
623 | 7. Soft limits |
624 | ||
625 | Soft limits allow for greater sharing of memory. The idea behind soft limits | |
626 | is to allow control groups to use as much of the memory as needed, provided | |
627 | ||
628 | a. There is no memory contention | |
629 | b. They do not exceed their hard limit | |
630 | ||
dc10e281 | 631 | When the system detects memory contention or low memory, control groups |
a6df6361 BS |
632 | are pushed back to their soft limits. If the soft limit of each control |
633 | group is very high, they are pushed back as much as possible to make | |
634 | sure that one control group does not starve the others of memory. | |
635 | ||
1939c557 | 636 | Please note that soft limits is a best-effort feature; it comes with |
a6df6361 BS |
637 | no guarantees, but it does its best to make sure that when memory is |
638 | heavily contended for, memory is allocated based on the soft limit | |
1939c557 | 639 | hints/setup. Currently soft limit based reclaim is set up such that |
a6df6361 BS |
640 | it gets invoked from balance_pgdat (kswapd). |
641 | ||
642 | 7.1 Interface | |
643 | ||
644 | Soft limits can be setup by using the following commands (in this example we | |
dc10e281 | 645 | assume a soft limit of 256 MiB) |
a6df6361 BS |
646 | |
647 | # echo 256M > memory.soft_limit_in_bytes | |
648 | ||
649 | If we want to change this to 1G, we can at any time use | |
650 | ||
651 | # echo 1G > memory.soft_limit_in_bytes | |
652 | ||
653 | NOTE1: Soft limits take effect over a long period of time, since they involve | |
654 | reclaiming memory for balancing between memory cgroups | |
655 | NOTE2: It is recommended to set the soft limit always below the hard limit, | |
656 | otherwise the hard limit will take precedence. | |
657 | ||
7dc74be0 DN |
658 | 8. Move charges at task migration |
659 | ||
660 | Users can move charges associated with a task along with task migration, that | |
661 | is, uncharge task's pages from the old cgroup and charge them to the new cgroup. | |
02491447 DN |
662 | This feature is not supported in !CONFIG_MMU environments because of lack of |
663 | page tables. | |
7dc74be0 DN |
664 | |
665 | 8.1 Interface | |
666 | ||
1939c557 | 667 | This feature is disabled by default. It can be enabledi (and disabled again) by |
7dc74be0 DN |
668 | writing to memory.move_charge_at_immigrate of the destination cgroup. |
669 | ||
670 | If you want to enable it: | |
671 | ||
672 | # echo (some positive value) > memory.move_charge_at_immigrate | |
673 | ||
674 | Note: Each bits of move_charge_at_immigrate has its own meaning about what type | |
675 | of charges should be moved. See 8.2 for details. | |
1939c557 MK |
676 | Note: Charges are moved only when you move mm->owner, in other words, |
677 | a leader of a thread group. | |
7dc74be0 DN |
678 | Note: If we cannot find enough space for the task in the destination cgroup, we |
679 | try to make space by reclaiming memory. Task migration may fail if we | |
680 | cannot make enough space. | |
dc10e281 | 681 | Note: It can take several seconds if you move charges much. |
7dc74be0 DN |
682 | |
683 | And if you want disable it again: | |
684 | ||
685 | # echo 0 > memory.move_charge_at_immigrate | |
686 | ||
1939c557 | 687 | 8.2 Type of charges which can be moved |
7dc74be0 | 688 | |
1939c557 MK |
689 | Each bit in move_charge_at_immigrate has its own meaning about what type of |
690 | charges should be moved. But in any case, it must be noted that an account of | |
691 | a page or a swap can be moved only when it is charged to the task's current | |
692 | (old) memory cgroup. | |
7dc74be0 DN |
693 | |
694 | bit | what type of charges would be moved ? | |
695 | -----+------------------------------------------------------------------------ | |
1939c557 MK |
696 | 0 | A charge of an anonymous page (or swap of it) used by the target task. |
697 | | You must enable Swap Extension (see 2.4) to enable move of swap charges. | |
87946a72 | 698 | -----+------------------------------------------------------------------------ |
1939c557 | 699 | 1 | A charge of file pages (normal file, tmpfs file (e.g. ipc shared memory) |
dc10e281 | 700 | | and swaps of tmpfs file) mmapped by the target task. Unlike the case of |
1939c557 | 701 | | anonymous pages, file pages (and swaps) in the range mmapped by the task |
87946a72 DN |
702 | | will be moved even if the task hasn't done page fault, i.e. they might |
703 | | not be the task's "RSS", but other task's "RSS" that maps the same file. | |
1939c557 MK |
704 | | And mapcount of the page is ignored (the page can be moved even if |
705 | | page_mapcount(page) > 1). You must enable Swap Extension (see 2.4) to | |
87946a72 | 706 | | enable move of swap charges. |
7dc74be0 DN |
707 | |
708 | 8.3 TODO | |
709 | ||
7dc74be0 DN |
710 | - All of moving charge operations are done under cgroup_mutex. It's not good |
711 | behavior to hold the mutex too long, so we may need some trick. | |
712 | ||
2e72b634 KS |
713 | 9. Memory thresholds |
714 | ||
1939c557 | 715 | Memory cgroup implements memory thresholds using the cgroups notification |
2e72b634 KS |
716 | API (see cgroups.txt). It allows to register multiple memory and memsw |
717 | thresholds and gets notifications when it crosses. | |
718 | ||
1939c557 | 719 | To register a threshold, an application must: |
dc10e281 KH |
720 | - create an eventfd using eventfd(2); |
721 | - open memory.usage_in_bytes or memory.memsw.usage_in_bytes; | |
722 | - write string like "<event_fd> <fd of memory.usage_in_bytes> <threshold>" to | |
723 | cgroup.event_control. | |
2e72b634 KS |
724 | |
725 | Application will be notified through eventfd when memory usage crosses | |
726 | threshold in any direction. | |
727 | ||
728 | It's applicable for root and non-root cgroup. | |
729 | ||
9490ff27 KH |
730 | 10. OOM Control |
731 | ||
3c11ecf4 KH |
732 | memory.oom_control file is for OOM notification and other controls. |
733 | ||
1939c557 | 734 | Memory cgroup implements OOM notifier using the cgroup notification |
dc10e281 KH |
735 | API (See cgroups.txt). It allows to register multiple OOM notification |
736 | delivery and gets notification when OOM happens. | |
9490ff27 | 737 | |
1939c557 | 738 | To register a notifier, an application must: |
9490ff27 KH |
739 | - create an eventfd using eventfd(2) |
740 | - open memory.oom_control file | |
dc10e281 KH |
741 | - write string like "<event_fd> <fd of memory.oom_control>" to |
742 | cgroup.event_control | |
9490ff27 | 743 | |
1939c557 MK |
744 | The application will be notified through eventfd when OOM happens. |
745 | OOM notification doesn't work for the root cgroup. | |
9490ff27 | 746 | |
1939c557 | 747 | You can disable the OOM-killer by writing "1" to memory.oom_control file, as: |
dc10e281 | 748 | |
3c11ecf4 KH |
749 | #echo 1 > memory.oom_control |
750 | ||
1939c557 | 751 | This operation is only allowed to the top cgroup of a sub-hierarchy. |
dc10e281 KH |
752 | If OOM-killer is disabled, tasks under cgroup will hang/sleep |
753 | in memory cgroup's OOM-waitqueue when they request accountable memory. | |
3c11ecf4 | 754 | |
dc10e281 | 755 | For running them, you have to relax the memory cgroup's OOM status by |
3c11ecf4 KH |
756 | * enlarge limit or reduce usage. |
757 | To reduce usage, | |
758 | * kill some tasks. | |
759 | * move some tasks to other group with account migration. | |
760 | * remove some files (on tmpfs?) | |
761 | ||
762 | Then, stopped tasks will work again. | |
763 | ||
764 | At reading, current status of OOM is shown. | |
765 | oom_kill_disable 0 or 1 (if 1, oom-killer is disabled) | |
dc10e281 | 766 | under_oom 0 or 1 (if 1, the memory cgroup is under OOM, tasks may |
3c11ecf4 | 767 | be stopped.) |
9490ff27 | 768 | |
70ddf637 AV |
769 | 11. Memory Pressure |
770 | ||
771 | The pressure level notifications can be used to monitor the memory | |
772 | allocation cost; based on the pressure, applications can implement | |
773 | different strategies of managing their memory resources. The pressure | |
774 | levels are defined as following: | |
775 | ||
776 | The "low" level means that the system is reclaiming memory for new | |
777 | allocations. Monitoring this reclaiming activity might be useful for | |
778 | maintaining cache level. Upon notification, the program (typically | |
779 | "Activity Manager") might analyze vmstat and act in advance (i.e. | |
780 | prematurely shutdown unimportant services). | |
781 | ||
782 | The "medium" level means that the system is experiencing medium memory | |
783 | pressure, the system might be making swap, paging out active file caches, | |
784 | etc. Upon this event applications may decide to further analyze | |
785 | vmstat/zoneinfo/memcg or internal memory usage statistics and free any | |
786 | resources that can be easily reconstructed or re-read from a disk. | |
787 | ||
788 | The "critical" level means that the system is actively thrashing, it is | |
789 | about to out of memory (OOM) or even the in-kernel OOM killer is on its | |
790 | way to trigger. Applications should do whatever they can to help the | |
791 | system. It might be too late to consult with vmstat or any other | |
792 | statistics, so it's advisable to take an immediate action. | |
793 | ||
794 | The events are propagated upward until the event is handled, i.e. the | |
795 | events are not pass-through. Here is what this means: for example you have | |
796 | three cgroups: A->B->C. Now you set up an event listener on cgroups A, B | |
797 | and C, and suppose group C experiences some pressure. In this situation, | |
798 | only group C will receive the notification, i.e. groups A and B will not | |
799 | receive it. This is done to avoid excessive "broadcasting" of messages, | |
800 | which disturbs the system and which is especially bad if we are low on | |
801 | memory or thrashing. So, organize the cgroups wisely, or propagate the | |
802 | events manually (or, ask us to implement the pass-through events, | |
803 | explaining why would you need them.) | |
804 | ||
805 | The file memory.pressure_level is only used to setup an eventfd. To | |
806 | register a notification, an application must: | |
807 | ||
808 | - create an eventfd using eventfd(2); | |
809 | - open memory.pressure_level; | |
810 | - write string like "<event_fd> <fd of memory.pressure_level> <level>" | |
811 | to cgroup.event_control. | |
812 | ||
813 | Application will be notified through eventfd when memory pressure is at | |
814 | the specific level (or higher). Read/write operations to | |
815 | memory.pressure_level are no implemented. | |
816 | ||
817 | Test: | |
818 | ||
819 | Here is a small script example that makes a new cgroup, sets up a | |
820 | memory limit, sets up a notification in the cgroup and then makes child | |
821 | cgroup experience a critical pressure: | |
822 | ||
823 | # cd /sys/fs/cgroup/memory/ | |
824 | # mkdir foo | |
825 | # cd foo | |
826 | # cgroup_event_listener memory.pressure_level low & | |
827 | # echo 8000000 > memory.limit_in_bytes | |
828 | # echo 8000000 > memory.memsw.limit_in_bytes | |
829 | # echo $$ > tasks | |
830 | # dd if=/dev/zero | read x | |
831 | ||
832 | (Expect a bunch of notifications, and eventually, the oom-killer will | |
833 | trigger.) | |
834 | ||
835 | 12. TODO | |
1b6df3aa BS |
836 | |
837 | 1. Add support for accounting huge pages (as a separate controller) | |
dfc05c25 KH |
838 | 2. Make per-cgroup scanner reclaim not-shared pages first |
839 | 3. Teach controller to account for shared-pages | |
628f4235 | 840 | 4. Start reclamation in the background when the limit is |
1b6df3aa | 841 | not yet hit but the usage is getting closer |
1b6df3aa BS |
842 | |
843 | Summary | |
844 | ||
845 | Overall, the memory controller has been a stable controller and has been | |
846 | commented and discussed quite extensively in the community. | |
847 | ||
848 | References | |
849 | ||
850 | 1. Singh, Balbir. RFC: Memory Controller, http://lwn.net/Articles/206697/ | |
851 | 2. Singh, Balbir. Memory Controller (RSS Control), | |
852 | http://lwn.net/Articles/222762/ | |
853 | 3. Emelianov, Pavel. Resource controllers based on process cgroups | |
854 | http://lkml.org/lkml/2007/3/6/198 | |
855 | 4. Emelianov, Pavel. RSS controller based on process cgroups (v2) | |
2324c5dd | 856 | http://lkml.org/lkml/2007/4/9/78 |
1b6df3aa BS |
857 | 5. Emelianov, Pavel. RSS controller based on process cgroups (v3) |
858 | http://lkml.org/lkml/2007/5/30/244 | |
859 | 6. Menage, Paul. Control Groups v10, http://lwn.net/Articles/236032/ | |
860 | 7. Vaidyanathan, Srinivasan, Control Groups: Pagecache accounting and control | |
861 | subsystem (v3), http://lwn.net/Articles/235534/ | |
2324c5dd | 862 | 8. Singh, Balbir. RSS controller v2 test results (lmbench), |
1b6df3aa | 863 | http://lkml.org/lkml/2007/5/17/232 |
2324c5dd | 864 | 9. Singh, Balbir. RSS controller v2 AIM9 results |
1b6df3aa | 865 | http://lkml.org/lkml/2007/5/18/1 |
2324c5dd | 866 | 10. Singh, Balbir. Memory controller v6 test results, |
1b6df3aa | 867 | http://lkml.org/lkml/2007/8/19/36 |
2324c5dd LZ |
868 | 11. Singh, Balbir. Memory controller introduction (v6), |
869 | http://lkml.org/lkml/2007/8/17/69 | |
1b6df3aa BS |
870 | 12. Corbet, Jonathan, Controlling memory use in cgroups, |
871 | http://lwn.net/Articles/243795/ |