1 Documentation for /proc/sys/vm/* kernel version 2.2.10
2 (c) 1998, 1999, Rik van Riel <riel@nl.linux.org>
4 For general info and legal blurb, please look in README.
6 ==============================================================
8 This file contains the documentation for the sysctl files in
9 /proc/sys/vm and is valid for Linux kernel version 2.2.
11 The files in this directory can be used to tune the operation
12 of the virtual memory (VM) subsystem of the Linux kernel and
13 the writeout of dirty data to disk.
15 Default values and initialization routines for most of these
16 files can be found in mm/swap.c.
18 Currently, these files are in /proc/sys/vm:
22 - dirty_background_ratio
23 - dirty_expire_centisecs
24 - dirty_writeback_centisecs
34 ==============================================================
36 dirty_ratio, dirty_background_ratio, dirty_expire_centisecs,
37 dirty_writeback_centisecs, vfs_cache_pressure, laptop_mode,
38 block_dump, swap_token_timeout, drop-caches:
40 See Documentation/filesystems/proc.txt
42 ==============================================================
46 This value contains a flag that enables memory overcommitment.
48 When this flag is 0, the kernel attempts to estimate the amount
49 of free memory left when userspace requests more memory.
51 When this flag is 1, the kernel pretends there is always enough
52 memory until it actually runs out.
54 When this flag is 2, the kernel uses a "never overcommit"
55 policy that attempts to prevent any overcommit of memory.
57 This feature can be very useful because there are a lot of
58 programs that malloc() huge amounts of memory "just-in-case"
59 and don't use much of it.
61 The default value is 0.
63 See Documentation/vm/overcommit-accounting and
64 security/commoncap.c::cap_vm_enough_memory() for more information.
66 ==============================================================
70 When overcommit_memory is set to 2, the committed address
71 space is not permitted to exceed swap plus this percentage
72 of physical RAM. See above.
74 ==============================================================
78 The Linux VM subsystem avoids excessive disk seeks by reading
79 multiple pages on a page fault. The number of pages it reads
80 is dependent on the amount of memory in your machine.
82 The number of pages the kernel reads in at once is equal to
83 2 ^ page-cluster. Values above 2 ^ 5 don't make much sense
84 for swap because we only cluster swap data in 32-page groups.
86 ==============================================================
90 This file contains the maximum number of memory map areas a process
91 may have. Memory map areas are used as a side-effect of calling
92 malloc, directly by mmap and mprotect, and also when loading shared
95 While most applications need less than a thousand maps, certain
96 programs, particularly malloc debuggers, may consume lots of them,
97 e.g., up to one or two maps per allocation.
99 The default value is 65536.
101 ==============================================================
105 This is used to force the Linux VM to keep a minimum number
106 of kilobytes free. The VM uses this number to compute a pages_min
107 value for each lowmem zone in the system. Each lowmem zone gets
108 a number of reserved free pages based proportionally on its size.
110 ==============================================================
112 percpu_pagelist_fraction
114 This is the fraction of pages at most (high mark pcp->high) in each zone that
115 are allocated for each per cpu page list. The min value for this is 8. It
116 means that we don't allow more than 1/8th of pages in each zone to be
117 allocated in any single per_cpu_pagelist. This entry only changes the value
118 of hot per cpu pagelists. User can specify a number like 100 to allocate
119 1/100th of each zone to each per cpu page list.
121 The batch value of each per cpu pagelist is also updated as a result. It is
122 set to pcp->high/4. The upper limit of batch is (PAGE_SHIFT * 8)
124 The initial value is zero. Kernel does not use this value at boot time to set
125 the high water marks for each per cpu page list.
127 ===============================================================
131 Zone_reclaim_mode allows to set more or less agressive approaches to
132 reclaim memory when a zone runs out of memory. If it is set to zero then no
133 zone reclaim occurs. Allocations will be satisfied from other zones / nodes
136 This is value ORed together of
139 2 = Zone reclaim writes dirty pages out
140 4 = Zone reclaim swaps pages
141 8 = Also do a global slab reclaim pass
143 zone_reclaim_mode is set during bootup to 1 if it is determined that pages
144 from remote zones will cause a measurable performance reduction. The
145 page allocator will then reclaim easily reusable pages (those page
146 cache pages that are currently not used) before allocating off node pages.
148 It may be beneficial to switch off zone reclaim if the system is
149 used for a file server and all of memory should be used for caching files
150 from disk. In that case the caching effect is more important than
153 Allowing zone reclaim to write out pages stops processes that are
154 writing large amounts of data from dirtying pages on other nodes. Zone
155 reclaim will write out dirty pages if a zone fills up and so effectively
156 throttle the process. This may decrease the performance of a single process
157 since it cannot use all of system memory to buffer the outgoing writes
158 anymore but it preserve the memory on other nodes so that the performance
159 of other processes running on other nodes will not be affected.
161 Allowing regular swap effectively restricts allocations to the local
162 node unless explicitly overridden by memory policies or cpuset
165 It may be advisable to allow slab reclaim if the system makes heavy
166 use of files and builds up large slab caches. However, the slab
167 shrink operation is global, may take a long time and free slabs
168 in all nodes of the system.
170 =============================================================
174 This is available only on NUMA kernels.
176 A percentage of the file backed pages in each zone. Zone reclaim will only
177 occur if more than this percentage of pages are file backed and unmapped.
178 This is to insure that a minimal amount of local pages is still available for
179 file I/O even if the node is overallocated.
181 The default is 1 percent.
183 =============================================================
187 This enables or disables panic on out-of-memory feature. If this is set to 1,
188 the kernel panics when out-of-memory happens. If this is set to 0, the kernel
189 will kill some rogue process, called oom_killer. Usually, oom_killer can kill
190 rogue processes and system will survive. If you want to panic the system
191 rather than killing rogue processes, set this to 1.
193 The default value is 0.