Commit | Line | Data |
---|---|---|
db0fb184 | 1 | Documentation for /proc/sys/vm/* kernel version 2.6.29 |
1da177e4 | 2 | (c) 1998, 1999, Rik van Riel <riel@nl.linux.org> |
db0fb184 | 3 | (c) 2008 Peter W. Morreale <pmorreale@novell.com> |
1da177e4 LT |
4 | |
5 | For general info and legal blurb, please look in README. | |
6 | ||
7 | ============================================================== | |
8 | ||
9 | This file contains the documentation for the sysctl files in | |
db0fb184 | 10 | /proc/sys/vm and is valid for Linux kernel version 2.6.29. |
1da177e4 LT |
11 | |
12 | The files in this directory can be used to tune the operation | |
13 | of the virtual memory (VM) subsystem of the Linux kernel and | |
14 | the writeout of dirty data to disk. | |
15 | ||
16 | Default values and initialization routines for most of these | |
17 | files can be found in mm/swap.c. | |
18 | ||
19 | Currently, these files are in /proc/sys/vm: | |
db0fb184 | 20 | |
4eeab4f5 | 21 | - admin_reserve_kbytes |
db0fb184 | 22 | - block_dump |
76ab0f53 | 23 | - compact_memory |
5bbe3547 | 24 | - compact_unevictable_allowed |
db0fb184 | 25 | - dirty_background_bytes |
1da177e4 | 26 | - dirty_background_ratio |
db0fb184 | 27 | - dirty_bytes |
1da177e4 | 28 | - dirty_expire_centisecs |
db0fb184 | 29 | - dirty_ratio |
1da177e4 | 30 | - dirty_writeback_centisecs |
db0fb184 | 31 | - drop_caches |
5e771905 | 32 | - extfrag_threshold |
db0fb184 PM |
33 | - hugepages_treat_as_movable |
34 | - hugetlb_shm_group | |
35 | - laptop_mode | |
36 | - legacy_va_layout | |
37 | - lowmem_reserve_ratio | |
1da177e4 | 38 | - max_map_count |
6a46079c AK |
39 | - memory_failure_early_kill |
40 | - memory_failure_recovery | |
1da177e4 | 41 | - min_free_kbytes |
0ff38490 | 42 | - min_slab_ratio |
db0fb184 PM |
43 | - min_unmapped_ratio |
44 | - mmap_min_addr | |
d5dbac87 NA |
45 | - nr_hugepages |
46 | - nr_overcommit_hugepages | |
db0fb184 PM |
47 | - nr_trim_pages (only if CONFIG_MMU=n) |
48 | - numa_zonelist_order | |
49 | - oom_dump_tasks | |
50 | - oom_kill_allocating_task | |
49f0ce5f | 51 | - overcommit_kbytes |
db0fb184 PM |
52 | - overcommit_memory |
53 | - overcommit_ratio | |
54 | - page-cluster | |
55 | - panic_on_oom | |
56 | - percpu_pagelist_fraction | |
57 | - stat_interval | |
58 | - swappiness | |
c9b1d098 | 59 | - user_reserve_kbytes |
db0fb184 PM |
60 | - vfs_cache_pressure |
61 | - zone_reclaim_mode | |
62 | ||
1da177e4 LT |
63 | ============================================================== |
64 | ||
4eeab4f5 AS |
65 | admin_reserve_kbytes |
66 | ||
67 | The amount of free memory in the system that should be reserved for users | |
68 | with the capability cap_sys_admin. | |
69 | ||
70 | admin_reserve_kbytes defaults to min(3% of free pages, 8MB) | |
71 | ||
72 | That should provide enough for the admin to log in and kill a process, | |
73 | if necessary, under the default overcommit 'guess' mode. | |
74 | ||
75 | Systems running under overcommit 'never' should increase this to account | |
76 | for the full Virtual Memory Size of programs used to recover. Otherwise, | |
77 | root may not be able to log in to recover the system. | |
78 | ||
79 | How do you calculate a minimum useful reserve? | |
80 | ||
81 | sshd or login + bash (or some other shell) + top (or ps, kill, etc.) | |
82 | ||
83 | For overcommit 'guess', we can sum resident set sizes (RSS). | |
84 | On x86_64 this is about 8MB. | |
85 | ||
86 | For overcommit 'never', we can take the max of their virtual sizes (VSZ) | |
87 | and add the sum of their RSS. | |
88 | On x86_64 this is about 128MB. | |
89 | ||
90 | Changing this takes effect whenever an application requests memory. | |
91 | ||
92 | ============================================================== | |
93 | ||
db0fb184 | 94 | block_dump |
1da177e4 | 95 | |
db0fb184 PM |
96 | block_dump enables block I/O debugging when set to a nonzero value. More |
97 | information on block I/O debugging is in Documentation/laptops/laptop-mode.txt. | |
1da177e4 LT |
98 | |
99 | ============================================================== | |
100 | ||
76ab0f53 MG |
101 | compact_memory |
102 | ||
103 | Available only when CONFIG_COMPACTION is set. When 1 is written to the file, | |
104 | all zones are compacted such that free memory is available in contiguous | |
105 | blocks where possible. This can be important for example in the allocation of | |
106 | huge pages although processes will also directly compact memory as required. | |
107 | ||
108 | ============================================================== | |
109 | ||
5bbe3547 EM |
110 | compact_unevictable_allowed |
111 | ||
112 | Available only when CONFIG_COMPACTION is set. When set to 1, compaction is | |
113 | allowed to examine the unevictable lru (mlocked pages) for pages to compact. | |
114 | This should be used on systems where stalls for minor page faults are an | |
115 | acceptable trade for large contiguous free memory. Set to 0 to prevent | |
116 | compaction from moving pages that are unevictable. Default value is 1. | |
117 | ||
118 | ============================================================== | |
119 | ||
db0fb184 | 120 | dirty_background_bytes |
1da177e4 | 121 | |
6601fac8 AB |
122 | Contains the amount of dirty memory at which the background kernel |
123 | flusher threads will start writeback. | |
1da177e4 | 124 | |
abffc020 AR |
125 | Note: dirty_background_bytes is the counterpart of dirty_background_ratio. Only |
126 | one of them may be specified at a time. When one sysctl is written it is | |
127 | immediately taken into account to evaluate the dirty memory limits and the | |
128 | other appears as 0 when read. | |
1da177e4 | 129 | |
db0fb184 | 130 | ============================================================== |
1da177e4 | 131 | |
db0fb184 | 132 | dirty_background_ratio |
1da177e4 | 133 | |
715ea41e ZL |
134 | Contains, as a percentage of total available memory that contains free pages |
135 | and reclaimable pages, the number of pages at which the background kernel | |
136 | flusher threads will start writing out dirty data. | |
137 | ||
138 | The total avaiable memory is not equal to total system memory. | |
1da177e4 | 139 | |
db0fb184 | 140 | ============================================================== |
1da177e4 | 141 | |
db0fb184 PM |
142 | dirty_bytes |
143 | ||
144 | Contains the amount of dirty memory at which a process generating disk writes | |
145 | will itself start writeback. | |
146 | ||
abffc020 AR |
147 | Note: dirty_bytes is the counterpart of dirty_ratio. Only one of them may be |
148 | specified at a time. When one sysctl is written it is immediately taken into | |
149 | account to evaluate the dirty memory limits and the other appears as 0 when | |
150 | read. | |
1da177e4 | 151 | |
9e4a5bda AR |
152 | Note: the minimum value allowed for dirty_bytes is two pages (in bytes); any |
153 | value lower than this limit will be ignored and the old configuration will be | |
154 | retained. | |
155 | ||
1da177e4 LT |
156 | ============================================================== |
157 | ||
db0fb184 | 158 | dirty_expire_centisecs |
1da177e4 | 159 | |
db0fb184 | 160 | This tunable is used to define when dirty data is old enough to be eligible |
6601fac8 AB |
161 | for writeout by the kernel flusher threads. It is expressed in 100'ths |
162 | of a second. Data which has been dirty in-memory for longer than this | |
163 | interval will be written out next time a flusher thread wakes up. | |
db0fb184 PM |
164 | |
165 | ============================================================== | |
166 | ||
167 | dirty_ratio | |
168 | ||
715ea41e ZL |
169 | Contains, as a percentage of total available memory that contains free pages |
170 | and reclaimable pages, the number of pages at which a process which is | |
171 | generating disk writes will itself start writing out dirty data. | |
172 | ||
173 | The total avaiable memory is not equal to total system memory. | |
1da177e4 LT |
174 | |
175 | ============================================================== | |
176 | ||
db0fb184 | 177 | dirty_writeback_centisecs |
1da177e4 | 178 | |
6601fac8 | 179 | The kernel flusher threads will periodically wake up and write `old' data |
db0fb184 PM |
180 | out to disk. This tunable expresses the interval between those wakeups, in |
181 | 100'ths of a second. | |
1da177e4 | 182 | |
db0fb184 | 183 | Setting this to zero disables periodic writeback altogether. |
1da177e4 LT |
184 | |
185 | ============================================================== | |
186 | ||
db0fb184 | 187 | drop_caches |
1da177e4 | 188 | |
5509a5d2 DH |
189 | Writing to this will cause the kernel to drop clean caches, as well as |
190 | reclaimable slab objects like dentries and inodes. Once dropped, their | |
191 | memory becomes free. | |
1da177e4 | 192 | |
db0fb184 PM |
193 | To free pagecache: |
194 | echo 1 > /proc/sys/vm/drop_caches | |
5509a5d2 | 195 | To free reclaimable slab objects (includes dentries and inodes): |
db0fb184 | 196 | echo 2 > /proc/sys/vm/drop_caches |
5509a5d2 | 197 | To free slab objects and pagecache: |
db0fb184 | 198 | echo 3 > /proc/sys/vm/drop_caches |
1da177e4 | 199 | |
5509a5d2 DH |
200 | This is a non-destructive operation and will not free any dirty objects. |
201 | To increase the number of objects freed by this operation, the user may run | |
202 | `sync' prior to writing to /proc/sys/vm/drop_caches. This will minimize the | |
203 | number of dirty objects on the system and create more candidates to be | |
204 | dropped. | |
205 | ||
206 | This file is not a means to control the growth of the various kernel caches | |
207 | (inodes, dentries, pagecache, etc...) These objects are automatically | |
208 | reclaimed by the kernel when memory is needed elsewhere on the system. | |
209 | ||
210 | Use of this file can cause performance problems. Since it discards cached | |
211 | objects, it may cost a significant amount of I/O and CPU to recreate the | |
212 | dropped objects, especially if they were under heavy use. Because of this, | |
213 | use outside of a testing or debugging environment is not recommended. | |
214 | ||
215 | You may see informational messages in your kernel log when this file is | |
216 | used: | |
217 | ||
218 | cat (1234): drop_caches: 3 | |
219 | ||
220 | These are informational only. They do not mean that anything is wrong | |
221 | with your system. To disable them, echo 4 (bit 3) into drop_caches. | |
1da177e4 LT |
222 | |
223 | ============================================================== | |
224 | ||
5e771905 MG |
225 | extfrag_threshold |
226 | ||
227 | This parameter affects whether the kernel will compact memory or direct | |
228 | reclaim to satisfy a high-order allocation. /proc/extfrag_index shows what | |
229 | the fragmentation index for each order is in each zone in the system. Values | |
230 | tending towards 0 imply allocations would fail due to lack of memory, | |
231 | values towards 1000 imply failures are due to fragmentation and -1 implies | |
232 | that the allocation will succeed as long as watermarks are met. | |
233 | ||
234 | The kernel will not compact memory in a zone if the | |
235 | fragmentation index is <= extfrag_threshold. The default value is 500. | |
236 | ||
237 | ============================================================== | |
238 | ||
db0fb184 | 239 | hugepages_treat_as_movable |
1da177e4 | 240 | |
86cdb465 NH |
241 | This parameter controls whether we can allocate hugepages from ZONE_MOVABLE |
242 | or not. If set to non-zero, hugepages can be allocated from ZONE_MOVABLE. | |
243 | ZONE_MOVABLE is created when kernel boot parameter kernelcore= is specified, | |
244 | so this parameter has no effect if used without kernelcore=. | |
245 | ||
246 | Hugepage migration is now available in some situations which depend on the | |
247 | architecture and/or the hugepage size. If a hugepage supports migration, | |
248 | allocation from ZONE_MOVABLE is always enabled for the hugepage regardless | |
249 | of the value of this parameter. | |
250 | IOW, this parameter affects only non-migratable hugepages. | |
251 | ||
252 | Assuming that hugepages are not migratable in your system, one usecase of | |
253 | this parameter is that users can make hugepage pool more extensible by | |
254 | enabling the allocation from ZONE_MOVABLE. This is because on ZONE_MOVABLE | |
255 | page reclaim/migration/compaction work more and you can get contiguous | |
256 | memory more likely. Note that using ZONE_MOVABLE for non-migratable | |
257 | hugepages can do harm to other features like memory hotremove (because | |
258 | memory hotremove expects that memory blocks on ZONE_MOVABLE are always | |
259 | removable,) so it's a trade-off responsible for the users. | |
24950898 | 260 | |
8ad4b1fb RS |
261 | ============================================================== |
262 | ||
db0fb184 | 263 | hugetlb_shm_group |
8ad4b1fb | 264 | |
db0fb184 PM |
265 | hugetlb_shm_group contains group id that is allowed to create SysV |
266 | shared memory segment using hugetlb page. | |
8ad4b1fb | 267 | |
db0fb184 | 268 | ============================================================== |
8ad4b1fb | 269 | |
db0fb184 | 270 | laptop_mode |
1743660b | 271 | |
db0fb184 PM |
272 | laptop_mode is a knob that controls "laptop mode". All the things that are |
273 | controlled by this knob are discussed in Documentation/laptops/laptop-mode.txt. | |
1743660b | 274 | |
db0fb184 | 275 | ============================================================== |
1743660b | 276 | |
db0fb184 | 277 | legacy_va_layout |
1b2ffb78 | 278 | |
2174efb6 | 279 | If non-zero, this sysctl disables the new 32-bit mmap layout - the kernel |
db0fb184 | 280 | will use the legacy (2.4) layout for all processes. |
1b2ffb78 | 281 | |
db0fb184 | 282 | ============================================================== |
1b2ffb78 | 283 | |
db0fb184 PM |
284 | lowmem_reserve_ratio |
285 | ||
286 | For some specialised workloads on highmem machines it is dangerous for | |
287 | the kernel to allow process memory to be allocated from the "lowmem" | |
288 | zone. This is because that memory could then be pinned via the mlock() | |
289 | system call, or by unavailability of swapspace. | |
290 | ||
291 | And on large highmem machines this lack of reclaimable lowmem memory | |
292 | can be fatal. | |
293 | ||
294 | So the Linux page allocator has a mechanism which prevents allocations | |
295 | which _could_ use highmem from using too much lowmem. This means that | |
296 | a certain amount of lowmem is defended from the possibility of being | |
297 | captured into pinned user memory. | |
298 | ||
299 | (The same argument applies to the old 16 megabyte ISA DMA region. This | |
300 | mechanism will also defend that region from allocations which could use | |
301 | highmem or lowmem). | |
302 | ||
303 | The `lowmem_reserve_ratio' tunable determines how aggressive the kernel is | |
304 | in defending these lower zones. | |
305 | ||
306 | If you have a machine which uses highmem or ISA DMA and your | |
307 | applications are using mlock(), or if you are running with no swap then | |
308 | you probably should change the lowmem_reserve_ratio setting. | |
309 | ||
310 | The lowmem_reserve_ratio is an array. You can see them by reading this file. | |
311 | - | |
312 | % cat /proc/sys/vm/lowmem_reserve_ratio | |
313 | 256 256 32 | |
314 | - | |
315 | Note: # of this elements is one fewer than number of zones. Because the highest | |
316 | zone's value is not necessary for following calculation. | |
317 | ||
318 | But, these values are not used directly. The kernel calculates # of protection | |
319 | pages for each zones from them. These are shown as array of protection pages | |
320 | in /proc/zoneinfo like followings. (This is an example of x86-64 box). | |
321 | Each zone has an array of protection pages like this. | |
322 | ||
323 | - | |
324 | Node 0, zone DMA | |
325 | pages free 1355 | |
326 | min 3 | |
327 | low 3 | |
328 | high 4 | |
329 | : | |
330 | : | |
331 | numa_other 0 | |
332 | protection: (0, 2004, 2004, 2004) | |
333 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
334 | pagesets | |
335 | cpu: 0 pcp: 0 | |
336 | : | |
337 | - | |
338 | These protections are added to score to judge whether this zone should be used | |
339 | for page allocation or should be reclaimed. | |
340 | ||
341 | In this example, if normal pages (index=2) are required to this DMA zone and | |
41858966 MG |
342 | watermark[WMARK_HIGH] is used for watermark, the kernel judges this zone should |
343 | not be used because pages_free(1355) is smaller than watermark + protection[2] | |
db0fb184 PM |
344 | (4 + 2004 = 2008). If this protection value is 0, this zone would be used for |
345 | normal page requirement. If requirement is DMA zone(index=0), protection[0] | |
346 | (=0) is used. | |
347 | ||
348 | zone[i]'s protection[j] is calculated by following expression. | |
349 | ||
350 | (i < j): | |
351 | zone[i]->protection[j] | |
352 | = (total sums of present_pages from zone[i+1] to zone[j] on the node) | |
353 | / lowmem_reserve_ratio[i]; | |
354 | (i = j): | |
355 | (should not be protected. = 0; | |
356 | (i > j): | |
357 | (not necessary, but looks 0) | |
358 | ||
359 | The default values of lowmem_reserve_ratio[i] are | |
360 | 256 (if zone[i] means DMA or DMA32 zone) | |
361 | 32 (others). | |
362 | As above expression, they are reciprocal number of ratio. | |
363 | 256 means 1/256. # of protection pages becomes about "0.39%" of total present | |
364 | pages of higher zones on the node. | |
365 | ||
366 | If you would like to protect more pages, smaller values are effective. | |
367 | The minimum value is 1 (1/1 -> 100%). | |
1b2ffb78 | 368 | |
db0fb184 | 369 | ============================================================== |
1b2ffb78 | 370 | |
db0fb184 | 371 | max_map_count: |
1743660b | 372 | |
db0fb184 PM |
373 | This file contains the maximum number of memory map areas a process |
374 | may have. Memory map areas are used as a side-effect of calling | |
375 | malloc, directly by mmap and mprotect, and also when loading shared | |
376 | libraries. | |
1743660b | 377 | |
db0fb184 PM |
378 | While most applications need less than a thousand maps, certain |
379 | programs, particularly malloc debuggers, may consume lots of them, | |
380 | e.g., up to one or two maps per allocation. | |
fadd8fbd | 381 | |
db0fb184 | 382 | The default value is 65536. |
9614634f | 383 | |
6a46079c AK |
384 | ============================================================= |
385 | ||
386 | memory_failure_early_kill: | |
387 | ||
388 | Control how to kill processes when uncorrected memory error (typically | |
389 | a 2bit error in a memory module) is detected in the background by hardware | |
390 | that cannot be handled by the kernel. In some cases (like the page | |
391 | still having a valid copy on disk) the kernel will handle the failure | |
392 | transparently without affecting any applications. But if there is | |
393 | no other uptodate copy of the data it will kill to prevent any data | |
394 | corruptions from propagating. | |
395 | ||
396 | 1: Kill all processes that have the corrupted and not reloadable page mapped | |
397 | as soon as the corruption is detected. Note this is not supported | |
398 | for a few types of pages, like kernel internally allocated data or | |
399 | the swap cache, but works for the majority of user pages. | |
400 | ||
401 | 0: Only unmap the corrupted page from all processes and only kill a process | |
402 | who tries to access it. | |
403 | ||
404 | The kill is done using a catchable SIGBUS with BUS_MCEERR_AO, so processes can | |
405 | handle this if they want to. | |
406 | ||
407 | This is only active on architectures/platforms with advanced machine | |
408 | check handling and depends on the hardware capabilities. | |
409 | ||
410 | Applications can override this setting individually with the PR_MCE_KILL prctl | |
411 | ||
412 | ============================================================== | |
413 | ||
414 | memory_failure_recovery | |
415 | ||
416 | Enable memory failure recovery (when supported by the platform) | |
417 | ||
418 | 1: Attempt recovery. | |
419 | ||
420 | 0: Always panic on a memory failure. | |
421 | ||
db0fb184 | 422 | ============================================================== |
9614634f | 423 | |
db0fb184 | 424 | min_free_kbytes: |
9614634f | 425 | |
db0fb184 | 426 | This is used to force the Linux VM to keep a minimum number |
41858966 MG |
427 | of kilobytes free. The VM uses this number to compute a |
428 | watermark[WMARK_MIN] value for each lowmem zone in the system. | |
429 | Each lowmem zone gets a number of reserved free pages based | |
430 | proportionally on its size. | |
db0fb184 PM |
431 | |
432 | Some minimal amount of memory is needed to satisfy PF_MEMALLOC | |
433 | allocations; if you set this to lower than 1024KB, your system will | |
434 | become subtly broken, and prone to deadlock under high loads. | |
435 | ||
436 | Setting this too high will OOM your machine instantly. | |
9614634f CL |
437 | |
438 | ============================================================= | |
439 | ||
0ff38490 CL |
440 | min_slab_ratio: |
441 | ||
442 | This is available only on NUMA kernels. | |
443 | ||
444 | A percentage of the total pages in each zone. On Zone reclaim | |
445 | (fallback from the local zone occurs) slabs will be reclaimed if more | |
446 | than this percentage of pages in a zone are reclaimable slab pages. | |
447 | This insures that the slab growth stays under control even in NUMA | |
448 | systems that rarely perform global reclaim. | |
449 | ||
450 | The default is 5 percent. | |
451 | ||
452 | Note that slab reclaim is triggered in a per zone / node fashion. | |
453 | The process of reclaiming slab memory is currently not node specific | |
454 | and may not be fast. | |
455 | ||
456 | ============================================================= | |
457 | ||
db0fb184 | 458 | min_unmapped_ratio: |
fadd8fbd | 459 | |
db0fb184 | 460 | This is available only on NUMA kernels. |
fadd8fbd | 461 | |
90afa5de MG |
462 | This is a percentage of the total pages in each zone. Zone reclaim will |
463 | only occur if more than this percentage of pages are in a state that | |
464 | zone_reclaim_mode allows to be reclaimed. | |
465 | ||
466 | If zone_reclaim_mode has the value 4 OR'd, then the percentage is compared | |
467 | against all file-backed unmapped pages including swapcache pages and tmpfs | |
468 | files. Otherwise, only unmapped pages backed by normal files but not tmpfs | |
469 | files and similar are considered. | |
2b744c01 | 470 | |
db0fb184 | 471 | The default is 1 percent. |
fadd8fbd | 472 | |
db0fb184 | 473 | ============================================================== |
2b744c01 | 474 | |
db0fb184 | 475 | mmap_min_addr |
ed032189 | 476 | |
db0fb184 | 477 | This file indicates the amount of address space which a user process will |
af901ca1 | 478 | be restricted from mmapping. Since kernel null dereference bugs could |
db0fb184 PM |
479 | accidentally operate based on the information in the first couple of pages |
480 | of memory userspace processes should not be allowed to write to them. By | |
481 | default this value is set to 0 and no protections will be enforced by the | |
482 | security module. Setting this value to something like 64k will allow the | |
483 | vast majority of applications to work correctly and provide defense in depth | |
484 | against future potential kernel bugs. | |
fe071d7e | 485 | |
db0fb184 | 486 | ============================================================== |
fef1bdd6 | 487 | |
db0fb184 | 488 | nr_hugepages |
fef1bdd6 | 489 | |
db0fb184 | 490 | Change the minimum size of the hugepage pool. |
fef1bdd6 | 491 | |
db0fb184 | 492 | See Documentation/vm/hugetlbpage.txt |
fef1bdd6 | 493 | |
db0fb184 | 494 | ============================================================== |
fef1bdd6 | 495 | |
db0fb184 | 496 | nr_overcommit_hugepages |
fef1bdd6 | 497 | |
db0fb184 PM |
498 | Change the maximum size of the hugepage pool. The maximum is |
499 | nr_hugepages + nr_overcommit_hugepages. | |
fe071d7e | 500 | |
db0fb184 | 501 | See Documentation/vm/hugetlbpage.txt |
fe071d7e | 502 | |
db0fb184 | 503 | ============================================================== |
fe071d7e | 504 | |
db0fb184 | 505 | nr_trim_pages |
ed032189 | 506 | |
db0fb184 PM |
507 | This is available only on NOMMU kernels. |
508 | ||
509 | This value adjusts the excess page trimming behaviour of power-of-2 aligned | |
510 | NOMMU mmap allocations. | |
511 | ||
512 | A value of 0 disables trimming of allocations entirely, while a value of 1 | |
513 | trims excess pages aggressively. Any value >= 1 acts as the watermark where | |
514 | trimming of allocations is initiated. | |
515 | ||
516 | The default value is 1. | |
517 | ||
518 | See Documentation/nommu-mmap.txt for more information. | |
ed032189 | 519 | |
f0c0b2b8 KH |
520 | ============================================================== |
521 | ||
522 | numa_zonelist_order | |
523 | ||
524 | This sysctl is only for NUMA. | |
525 | 'where the memory is allocated from' is controlled by zonelists. | |
526 | (This documentation ignores ZONE_HIGHMEM/ZONE_DMA32 for simple explanation. | |
527 | you may be able to read ZONE_DMA as ZONE_DMA32...) | |
528 | ||
529 | In non-NUMA case, a zonelist for GFP_KERNEL is ordered as following. | |
530 | ZONE_NORMAL -> ZONE_DMA | |
531 | This means that a memory allocation request for GFP_KERNEL will | |
532 | get memory from ZONE_DMA only when ZONE_NORMAL is not available. | |
533 | ||
534 | In NUMA case, you can think of following 2 types of order. | |
535 | Assume 2 node NUMA and below is zonelist of Node(0)'s GFP_KERNEL | |
536 | ||
537 | (A) Node(0) ZONE_NORMAL -> Node(0) ZONE_DMA -> Node(1) ZONE_NORMAL | |
538 | (B) Node(0) ZONE_NORMAL -> Node(1) ZONE_NORMAL -> Node(0) ZONE_DMA. | |
539 | ||
540 | Type(A) offers the best locality for processes on Node(0), but ZONE_DMA | |
541 | will be used before ZONE_NORMAL exhaustion. This increases possibility of | |
542 | out-of-memory(OOM) of ZONE_DMA because ZONE_DMA is tend to be small. | |
543 | ||
544 | Type(B) cannot offer the best locality but is more robust against OOM of | |
545 | the DMA zone. | |
546 | ||
547 | Type(A) is called as "Node" order. Type (B) is "Zone" order. | |
548 | ||
549 | "Node order" orders the zonelists by node, then by zone within each node. | |
5a3016a6 | 550 | Specify "[Nn]ode" for node order |
f0c0b2b8 KH |
551 | |
552 | "Zone Order" orders the zonelists by zone type, then by node within each | |
5a3016a6 | 553 | zone. Specify "[Zz]one" for zone order. |
f0c0b2b8 KH |
554 | |
555 | Specify "[Dd]efault" to request automatic configuration. Autoconfiguration | |
556 | will select "node" order in following case. | |
557 | (1) if the DMA zone does not exist or | |
558 | (2) if the DMA zone comprises greater than 50% of the available memory or | |
f8f191f1 | 559 | (3) if any node's DMA zone comprises greater than 70% of its local memory and |
f0c0b2b8 KH |
560 | the amount of local memory is big enough. |
561 | ||
562 | Otherwise, "zone" order will be selected. Default order is recommended unless | |
563 | this is causing problems for your system/application. | |
d5dbac87 NA |
564 | |
565 | ============================================================== | |
566 | ||
db0fb184 | 567 | oom_dump_tasks |
d5dbac87 | 568 | |
dc6c9a35 KS |
569 | Enables a system-wide task dump (excluding kernel threads) to be produced |
570 | when the kernel performs an OOM-killing and includes such information as | |
571 | pid, uid, tgid, vm size, rss, nr_ptes, nr_pmds, swapents, oom_score_adj | |
572 | score, and name. This is helpful to determine why the OOM killer was | |
573 | invoked, to identify the rogue task that caused it, and to determine why | |
574 | the OOM killer chose the task it did to kill. | |
d5dbac87 | 575 | |
db0fb184 PM |
576 | If this is set to zero, this information is suppressed. On very |
577 | large systems with thousands of tasks it may not be feasible to dump | |
578 | the memory state information for each one. Such systems should not | |
579 | be forced to incur a performance penalty in OOM conditions when the | |
580 | information may not be desired. | |
581 | ||
582 | If this is set to non-zero, this information is shown whenever the | |
583 | OOM killer actually kills a memory-hogging task. | |
584 | ||
ad915c43 | 585 | The default value is 1 (enabled). |
d5dbac87 NA |
586 | |
587 | ============================================================== | |
588 | ||
db0fb184 | 589 | oom_kill_allocating_task |
d5dbac87 | 590 | |
db0fb184 PM |
591 | This enables or disables killing the OOM-triggering task in |
592 | out-of-memory situations. | |
d5dbac87 | 593 | |
db0fb184 PM |
594 | If this is set to zero, the OOM killer will scan through the entire |
595 | tasklist and select a task based on heuristics to kill. This normally | |
596 | selects a rogue memory-hogging task that frees up a large amount of | |
597 | memory when killed. | |
598 | ||
599 | If this is set to non-zero, the OOM killer simply kills the task that | |
600 | triggered the out-of-memory condition. This avoids the expensive | |
601 | tasklist scan. | |
602 | ||
603 | If panic_on_oom is selected, it takes precedence over whatever value | |
604 | is used in oom_kill_allocating_task. | |
605 | ||
606 | The default value is 0. | |
dd8632a1 PM |
607 | |
608 | ============================================================== | |
609 | ||
49f0ce5f JM |
610 | overcommit_kbytes: |
611 | ||
612 | When overcommit_memory is set to 2, the committed address space is not | |
613 | permitted to exceed swap plus this amount of physical RAM. See below. | |
614 | ||
615 | Note: overcommit_kbytes is the counterpart of overcommit_ratio. Only one | |
616 | of them may be specified at a time. Setting one disables the other (which | |
617 | then appears as 0 when read). | |
618 | ||
619 | ============================================================== | |
620 | ||
db0fb184 | 621 | overcommit_memory: |
dd8632a1 | 622 | |
db0fb184 | 623 | This value contains a flag that enables memory overcommitment. |
dd8632a1 | 624 | |
db0fb184 PM |
625 | When this flag is 0, the kernel attempts to estimate the amount |
626 | of free memory left when userspace requests more memory. | |
dd8632a1 | 627 | |
db0fb184 PM |
628 | When this flag is 1, the kernel pretends there is always enough |
629 | memory until it actually runs out. | |
dd8632a1 | 630 | |
db0fb184 PM |
631 | When this flag is 2, the kernel uses a "never overcommit" |
632 | policy that attempts to prevent any overcommit of memory. | |
c9b1d098 | 633 | Note that user_reserve_kbytes affects this policy. |
dd8632a1 | 634 | |
db0fb184 PM |
635 | This feature can be very useful because there are a lot of |
636 | programs that malloc() huge amounts of memory "just-in-case" | |
637 | and don't use much of it. | |
638 | ||
639 | The default value is 0. | |
640 | ||
641 | See Documentation/vm/overcommit-accounting and | |
642 | security/commoncap.c::cap_vm_enough_memory() for more information. | |
643 | ||
644 | ============================================================== | |
645 | ||
646 | overcommit_ratio: | |
647 | ||
648 | When overcommit_memory is set to 2, the committed address | |
649 | space is not permitted to exceed swap plus this percentage | |
650 | of physical RAM. See above. | |
651 | ||
652 | ============================================================== | |
653 | ||
654 | page-cluster | |
655 | ||
df858fa8 CE |
656 | page-cluster controls the number of pages up to which consecutive pages |
657 | are read in from swap in a single attempt. This is the swap counterpart | |
658 | to page cache readahead. | |
659 | The mentioned consecutivity is not in terms of virtual/physical addresses, | |
660 | but consecutive on swap space - that means they were swapped out together. | |
db0fb184 PM |
661 | |
662 | It is a logarithmic value - setting it to zero means "1 page", setting | |
663 | it to 1 means "2 pages", setting it to 2 means "4 pages", etc. | |
df858fa8 | 664 | Zero disables swap readahead completely. |
db0fb184 PM |
665 | |
666 | The default value is three (eight pages at a time). There may be some | |
667 | small benefits in tuning this to a different value if your workload is | |
668 | swap-intensive. | |
669 | ||
df858fa8 CE |
670 | Lower values mean lower latencies for initial faults, but at the same time |
671 | extra faults and I/O delays for following faults if they would have been part of | |
672 | that consecutive pages readahead would have brought in. | |
673 | ||
db0fb184 PM |
674 | ============================================================= |
675 | ||
676 | panic_on_oom | |
677 | ||
678 | This enables or disables panic on out-of-memory feature. | |
679 | ||
680 | If this is set to 0, the kernel will kill some rogue process, | |
681 | called oom_killer. Usually, oom_killer can kill rogue processes and | |
682 | system will survive. | |
683 | ||
684 | If this is set to 1, the kernel panics when out-of-memory happens. | |
685 | However, if a process limits using nodes by mempolicy/cpusets, | |
686 | and those nodes become memory exhaustion status, one process | |
687 | may be killed by oom-killer. No panic occurs in this case. | |
688 | Because other nodes' memory may be free. This means system total status | |
689 | may be not fatal yet. | |
690 | ||
691 | If this is set to 2, the kernel panics compulsorily even on the | |
daaf1e68 KH |
692 | above-mentioned. Even oom happens under memory cgroup, the whole |
693 | system panics. | |
db0fb184 PM |
694 | |
695 | The default value is 0. | |
696 | 1 and 2 are for failover of clustering. Please select either | |
697 | according to your policy of failover. | |
daaf1e68 KH |
698 | panic_on_oom=2+kdump gives you very strong tool to investigate |
699 | why oom happens. You can get snapshot. | |
db0fb184 PM |
700 | |
701 | ============================================================= | |
702 | ||
703 | percpu_pagelist_fraction | |
704 | ||
705 | This is the fraction of pages at most (high mark pcp->high) in each zone that | |
706 | are allocated for each per cpu page list. The min value for this is 8. It | |
707 | means that we don't allow more than 1/8th of pages in each zone to be | |
708 | allocated in any single per_cpu_pagelist. This entry only changes the value | |
709 | of hot per cpu pagelists. User can specify a number like 100 to allocate | |
710 | 1/100th of each zone to each per cpu page list. | |
711 | ||
712 | The batch value of each per cpu pagelist is also updated as a result. It is | |
713 | set to pcp->high/4. The upper limit of batch is (PAGE_SHIFT * 8) | |
714 | ||
715 | The initial value is zero. Kernel does not use this value at boot time to set | |
7cd2b0a3 DR |
716 | the high water marks for each per cpu page list. If the user writes '0' to this |
717 | sysctl, it will revert to this default behavior. | |
db0fb184 PM |
718 | |
719 | ============================================================== | |
720 | ||
721 | stat_interval | |
722 | ||
723 | The time interval between which vm statistics are updated. The default | |
724 | is 1 second. | |
725 | ||
726 | ============================================================== | |
727 | ||
728 | swappiness | |
729 | ||
730 | This control is used to define how aggressive the kernel will swap | |
731 | memory pages. Higher values will increase agressiveness, lower values | |
8582cb96 AT |
732 | decrease the amount of swap. A value of 0 instructs the kernel not to |
733 | initiate swap until the amount of free and file-backed pages is less | |
734 | than the high water mark in a zone. | |
db0fb184 PM |
735 | |
736 | The default value is 60. | |
737 | ||
738 | ============================================================== | |
739 | ||
c9b1d098 AS |
740 | - user_reserve_kbytes |
741 | ||
633708a4 | 742 | When overcommit_memory is set to 2, "never overcommit" mode, reserve |
c9b1d098 AS |
743 | min(3% of current process size, user_reserve_kbytes) of free memory. |
744 | This is intended to prevent a user from starting a single memory hogging | |
745 | process, such that they cannot recover (kill the hog). | |
746 | ||
747 | user_reserve_kbytes defaults to min(3% of the current process size, 128MB). | |
748 | ||
749 | If this is reduced to zero, then the user will be allowed to allocate | |
750 | all free memory with a single process, minus admin_reserve_kbytes. | |
751 | Any subsequent attempts to execute a command will result in | |
752 | "fork: Cannot allocate memory". | |
753 | ||
754 | Changing this takes effect whenever an application requests memory. | |
755 | ||
756 | ============================================================== | |
757 | ||
db0fb184 PM |
758 | vfs_cache_pressure |
759 | ------------------ | |
760 | ||
4a0da71b DV |
761 | This percentage value controls the tendency of the kernel to reclaim |
762 | the memory which is used for caching of directory and inode objects. | |
db0fb184 PM |
763 | |
764 | At the default value of vfs_cache_pressure=100 the kernel will attempt to | |
765 | reclaim dentries and inodes at a "fair" rate with respect to pagecache and | |
766 | swapcache reclaim. Decreasing vfs_cache_pressure causes the kernel to prefer | |
55c37a84 JK |
767 | to retain dentry and inode caches. When vfs_cache_pressure=0, the kernel will |
768 | never reclaim dentries and inodes due to memory pressure and this can easily | |
769 | lead to out-of-memory conditions. Increasing vfs_cache_pressure beyond 100 | |
db0fb184 PM |
770 | causes the kernel to prefer to reclaim dentries and inodes. |
771 | ||
4a0da71b DV |
772 | Increasing vfs_cache_pressure significantly beyond 100 may have negative |
773 | performance impact. Reclaim code needs to take various locks to find freeable | |
774 | directory and inode objects. With vfs_cache_pressure=1000, it will look for | |
775 | ten times more freeable objects than there are. | |
776 | ||
db0fb184 PM |
777 | ============================================================== |
778 | ||
779 | zone_reclaim_mode: | |
780 | ||
781 | Zone_reclaim_mode allows someone to set more or less aggressive approaches to | |
782 | reclaim memory when a zone runs out of memory. If it is set to zero then no | |
783 | zone reclaim occurs. Allocations will be satisfied from other zones / nodes | |
784 | in the system. | |
785 | ||
786 | This is value ORed together of | |
787 | ||
788 | 1 = Zone reclaim on | |
789 | 2 = Zone reclaim writes dirty pages out | |
790 | 4 = Zone reclaim swaps pages | |
791 | ||
4f9b16a6 MG |
792 | zone_reclaim_mode is disabled by default. For file servers or workloads |
793 | that benefit from having their data cached, zone_reclaim_mode should be | |
794 | left disabled as the caching effect is likely to be more important than | |
db0fb184 PM |
795 | data locality. |
796 | ||
4f9b16a6 MG |
797 | zone_reclaim may be enabled if it's known that the workload is partitioned |
798 | such that each partition fits within a NUMA node and that accessing remote | |
799 | memory would cause a measurable performance reduction. The page allocator | |
800 | will then reclaim easily reusable pages (those page cache pages that are | |
801 | currently not used) before allocating off node pages. | |
802 | ||
db0fb184 PM |
803 | Allowing zone reclaim to write out pages stops processes that are |
804 | writing large amounts of data from dirtying pages on other nodes. Zone | |
805 | reclaim will write out dirty pages if a zone fills up and so effectively | |
806 | throttle the process. This may decrease the performance of a single process | |
807 | since it cannot use all of system memory to buffer the outgoing writes | |
808 | anymore but it preserve the memory on other nodes so that the performance | |
809 | of other processes running on other nodes will not be affected. | |
810 | ||
811 | Allowing regular swap effectively restricts allocations to the local | |
812 | node unless explicitly overridden by memory policies or cpuset | |
813 | configurations. | |
814 | ||
815 | ============ End of Document ================================= |