| blob | dbc3c6a3650f5c29b24000d69e1c9cb7973b4988 |
1 ------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3 ------------------------------------------------------------------------------
4 /proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
7 2.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8 ------------------------------------------------------------------------------
9 Version 1.3 Kernel version 2.2.12
10 Kernel version 2.4.0-test11-pre4
11 ------------------------------------------------------------------------------
13 Table of Contents
14 -----------------
16 0 Preface
17 0.1 Introduction/Credits
18 0.2 Legal Stuff
20 1 Collecting System Information
21 1.1 Process-Specific Subdirectories
22 1.2 Kernel data
23 1.3 IDE devices in /proc/ide
24 1.4 Networking info in /proc/net
25 1.5 SCSI info
26 1.6 Parallel port info in /proc/parport
27 1.7 TTY info in /proc/tty
28 1.8 Miscellaneous kernel statistics in /proc/stat
30 2 Modifying System Parameters
31 2.1 /proc/sys/fs - File system data
32 2.2 /proc/sys/fs/binfmt_misc - Miscellaneous binary formats
33 2.3 /proc/sys/kernel - general kernel parameters
34 2.4 /proc/sys/vm - The virtual memory subsystem
35 2.5 /proc/sys/dev - Device specific parameters
36 2.6 /proc/sys/sunrpc - Remote procedure calls
37 2.7 /proc/sys/net - Networking stuff
38 2.8 /proc/sys/net/ipv4 - IPV4 settings
39 2.9 Appletalk
40 2.10 IPX
41 2.11 /proc/sys/fs/mqueue - POSIX message queues filesystem
42 2.12 /proc/<pid>/oom_adj - Adjust the oom-killer score
43 2.13 /proc/<pid>/oom_score - Display current oom-killer score
44 2.14 /proc/<pid>/io - Display the IO accounting fields
45 2.15 /proc/<pid>/coredump_filter - Core dump filtering settings
46 2.16 /proc/<pid>/mountinfo - Information about mounts
48 ------------------------------------------------------------------------------
49 Preface
50 ------------------------------------------------------------------------------
52 0.1 Introduction/Credits
53 ------------------------
55 This documentation is part of a soon (or so we hope) to be released book on
56 the SuSE Linux distribution. As there is no complete documentation for the
57 /proc file system and we've used many freely available sources to write these
58 chapters, it seems only fair to give the work back to the Linux community.
59 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
60 afraid it's still far from complete, but we hope it will be useful. As far as
61 we know, it is the first 'all-in-one' document about the /proc file system. It
62 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
63 SPARC, AXP, etc., features, you probably won't find what you are looking for.
64 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
65 additions and patches are welcome and will be added to this document if you
66 mail them to Bodo.
68 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
69 other people for help compiling this documentation. We'd also like to extend a
70 special thank you to Andi Kleen for documentation, which we relied on heavily
71 to create this document, as well as the additional information he provided.
72 Thanks to everybody else who contributed source or docs to the Linux kernel
73 and helped create a great piece of software... :)
75 If you have any comments, corrections or additions, please don't hesitate to
76 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
77 document.
79 The latest version of this document is available online at
80 http://skaro.nightcrawler.com/~bb/Docs/Proc as HTML version.
82 If the above direction does not works for you, ypu could try the kernel
83 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
84 comandante@zaralinux.com.
86 0.2 Legal Stuff
87 ---------------
89 We don't guarantee the correctness of this document, and if you come to us
90 complaining about how you screwed up your system because of incorrect
91 documentation, we won't feel responsible...
93 ------------------------------------------------------------------------------
94 CHAPTER 1: COLLECTING SYSTEM INFORMATION
95 ------------------------------------------------------------------------------
97 ------------------------------------------------------------------------------
98 In This Chapter
99 ------------------------------------------------------------------------------
100 * Investigating the properties of the pseudo file system /proc and its
101 ability to provide information on the running Linux system
102 * Examining /proc's structure
103 * Uncovering various information about the kernel and the processes running
104 on the system
105 ------------------------------------------------------------------------------
108 The proc file system acts as an interface to internal data structures in the
109 kernel. It can be used to obtain information about the system and to change
110 certain kernel parameters at runtime (sysctl).
112 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
113 show you how you can use /proc/sys to change settings.
115 1.1 Process-Specific Subdirectories
116 -----------------------------------
118 The directory /proc contains (among other things) one subdirectory for each
119 process running on the system, which is named after the process ID (PID).
121 The link self points to the process reading the file system. Each process
122 subdirectory has the entries listed in Table 1-1.
125 Table 1-1: Process specific entries in /proc
126 ..............................................................................
127 File Content
128 clear_refs Clears page referenced bits shown in smaps output
129 cmdline Command line arguments
130 cpu Current and last cpu in which it was executed (2.4)(smp)
131 cwd Link to the current working directory
132 environ Values of environment variables
133 exe Link to the executable of this process
134 fd Directory, which contains all file descriptors
135 maps Memory maps to executables and library files (2.4)
136 mem Memory held by this process
137 root Link to the root directory of this process
138 stat Process status
139 statm Process memory status information
140 status Process status in human readable form
141 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
142 smaps Extension based on maps, the rss size for each mapped file
143 ..............................................................................
145 For example, to get the status information of a process, all you have to do is
146 read the file /proc/PID/status:
148 >cat /proc/self/status
149 Name: cat
150 State: R (running)
151 Pid: 5452
152 PPid: 743
153 TracerPid: 0 (2.4)
154 Uid: 501 501 501 501
155 Gid: 100 100 100 100
156 Groups: 100 14 16
157 VmSize: 1112 kB
158 VmLck: 0 kB
159 VmRSS: 348 kB
160 VmData: 24 kB
161 VmStk: 12 kB
162 VmExe: 8 kB
163 VmLib: 1044 kB
164 SigPnd: 0000000000000000
165 SigBlk: 0000000000000000
166 SigIgn: 0000000000000000
167 SigCgt: 0000000000000000
168 CapInh: 00000000fffffeff
169 CapPrm: 0000000000000000
170 CapEff: 0000000000000000
173 This shows you nearly the same information you would get if you viewed it with
174 the ps command. In fact, ps uses the proc file system to obtain its
175 information. The statm file contains more detailed information about the
176 process memory usage. Its seven fields are explained in Table 1-2. The stat
177 file contains details information about the process itself. Its fields are
178 explained in Table 1-3.
181 Table 1-2: Contents of the statm files (as of 2.6.8-rc3)
182 ..............................................................................
183 Field Content
184 size total program size (pages) (same as VmSize in status)
185 resident size of memory portions (pages) (same as VmRSS in status)
186 shared number of pages that are shared (i.e. backed by a file)
187 trs number of pages that are 'code' (not including libs; broken,
188 includes data segment)
189 lrs number of pages of library (always 0 on 2.6)
190 drs number of pages of data/stack (including libs; broken,
191 includes library text)
192 dt number of dirty pages (always 0 on 2.6)
193 ..............................................................................
196 Table 1-3: Contents of the stat files (as of 2.6.22-rc3)
197 ..............................................................................
198 Field Content
199 pid process id
200 tcomm filename of the executable
201 state state (R is running, S is sleeping, D is sleeping in an
202 uninterruptible wait, Z is zombie, T is traced or stopped)
203 ppid process id of the parent process
204 pgrp pgrp of the process
205 sid session id
206 tty_nr tty the process uses
207 tty_pgrp pgrp of the tty
208 flags task flags
209 min_flt number of minor faults
210 cmin_flt number of minor faults with child's
211 maj_flt number of major faults
212 cmaj_flt number of major faults with child's
213 utime user mode jiffies
214 stime kernel mode jiffies
215 cutime user mode jiffies with child's
216 cstime kernel mode jiffies with child's
217 priority priority level
218 nice nice level
219 num_threads number of threads
220 it_real_value (obsolete, always 0)
221 start_time time the process started after system boot
222 vsize virtual memory size
223 rss resident set memory size
224 rsslim current limit in bytes on the rss
225 start_code address above which program text can run
226 end_code address below which program text can run
227 start_stack address of the start of the stack
228 esp current value of ESP
229 eip current value of EIP
230 pending bitmap of pending signals (obsolete)
231 blocked bitmap of blocked signals (obsolete)
232 sigign bitmap of ignored signals (obsolete)
233 sigcatch bitmap of catched signals (obsolete)
234 wchan address where process went to sleep
235 0 (place holder)
236 0 (place holder)
237 exit_signal signal to send to parent thread on exit
238 task_cpu which CPU the task is scheduled on
239 rt_priority realtime priority
240 policy scheduling policy (man sched_setscheduler)
241 blkio_ticks time spent waiting for block IO
242 ..............................................................................
245 1.2 Kernel data
246 ---------------
248 Similar to the process entries, the kernel data files give information about
249 the running kernel. The files used to obtain this information are contained in
250 /proc and are listed in Table 1-4. Not all of these will be present in your
251 system. It depends on the kernel configuration and the loaded modules, which
252 files are there, and which are missing.
254 Table 1-4: Kernel info in /proc
255 ..............................................................................
256 File Content
257 apm Advanced power management info
258 buddyinfo Kernel memory allocator information (see text) (2.5)
259 bus Directory containing bus specific information
260 cmdline Kernel command line
261 cpuinfo Info about the CPU
262 devices Available devices (block and character)
263 dma Used DMS channels
264 filesystems Supported filesystems
265 driver Various drivers grouped here, currently rtc (2.4)
266 execdomains Execdomains, related to security (2.4)
267 fb Frame Buffer devices (2.4)
268 fs File system parameters, currently nfs/exports (2.4)
269 ide Directory containing info about the IDE subsystem
270 interrupts Interrupt usage
271 iomem Memory map (2.4)
272 ioports I/O port usage
273 irq Masks for irq to cpu affinity (2.4)(smp?)
274 isapnp ISA PnP (Plug&Play) Info (2.4)
275 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
276 kmsg Kernel messages
277 ksyms Kernel symbol table
278 loadavg Load average of last 1, 5 & 15 minutes
279 locks Kernel locks
280 meminfo Memory info
281 misc Miscellaneous
282 modules List of loaded modules
283 mounts Mounted filesystems
284 net Networking info (see text)
285 partitions Table of partitions known to the system
286 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
287 decoupled by lspci (2.4)
288 rtc Real time clock
289 scsi SCSI info (see text)
290 slabinfo Slab pool info
291 stat Overall statistics
292 swaps Swap space utilization
293 sys See chapter 2
294 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
295 tty Info of tty drivers
296 uptime System uptime
297 version Kernel version
298 video bttv info of video resources (2.4)
299 ..............................................................................
301 You can, for example, check which interrupts are currently in use and what
302 they are used for by looking in the file /proc/interrupts:
304 > cat /proc/interrupts
305 CPU0
306 0: 8728810 XT-PIC timer
307 1: 895 XT-PIC keyboard
308 2: 0 XT-PIC cascade
309 3: 531695 XT-PIC aha152x
310 4: 2014133 XT-PIC serial
311 5: 44401 XT-PIC pcnet_cs
312 8: 2 XT-PIC rtc
313 11: 8 XT-PIC i82365
314 12: 182918 XT-PIC PS/2 Mouse
315 13: 1 XT-PIC fpu
316 14: 1232265 XT-PIC ide0
317 15: 7 XT-PIC ide1
318 NMI: 0
320 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
321 output of a SMP machine):
323 > cat /proc/interrupts
325 CPU0 CPU1
326 0: 1243498 1214548 IO-APIC-edge timer
327 1: 8949 8958 IO-APIC-edge keyboard
328 2: 0 0 XT-PIC cascade
329 5: 11286 10161 IO-APIC-edge soundblaster
330 8: 1 0 IO-APIC-edge rtc
331 9: 27422 27407 IO-APIC-edge 3c503
332 12: 113645 113873 IO-APIC-edge PS/2 Mouse
333 13: 0 0 XT-PIC fpu
334 14: 22491 24012 IO-APIC-edge ide0
335 15: 2183 2415 IO-APIC-edge ide1
336 17: 30564 30414 IO-APIC-level eth0
337 18: 177 164 IO-APIC-level bttv
338 NMI: 2457961 2457959
339 LOC: 2457882 2457881
340 ERR: 2155
342 NMI is incremented in this case because every timer interrupt generates a NMI
343 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
345 LOC is the local interrupt counter of the internal APIC of every CPU.
347 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
348 connects the CPUs in a SMP system. This means that an error has been detected,
349 the IO-APIC automatically retry the transmission, so it should not be a big
350 problem, but you should read the SMP-FAQ.
352 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
353 /proc/interrupts to display every IRQ vector in use by the system, not
354 just those considered 'most important'. The new vectors are:
356 THR -- interrupt raised when a machine check threshold counter
357 (typically counting ECC corrected errors of memory or cache) exceeds
358 a configurable threshold. Only available on some systems.
360 TRM -- a thermal event interrupt occurs when a temperature threshold
361 has been exceeded for the CPU. This interrupt may also be generated
362 when the temperature drops back to normal.
364 SPU -- a spurious interrupt is some interrupt that was raised then lowered
365 by some IO device before it could be fully processed by the APIC. Hence
366 the APIC sees the interrupt but does not know what device it came from.
367 For this case the APIC will generate the interrupt with a IRQ vector
368 of 0xff. This might also be generated by chipset bugs.
370 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
371 sent from one CPU to another per the needs of the OS. Typically,
372 their statistics are used by kernel developers and interested users to
373 determine the occurance of interrupt of the given type.
375 The above IRQ vectors are displayed only when relevent. For example,
376 the threshold vector does not exist on x86_64 platforms. Others are
377 suppressed when the system is a uniprocessor. As of this writing, only
378 i386 and x86_64 platforms support the new IRQ vector displays.
380 Of some interest is the introduction of the /proc/irq directory to 2.4.
381 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
382 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
383 irq subdir is one subdir for each IRQ, and one file; prof_cpu_mask
385 For example
386 > ls /proc/irq/
387 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
388 1 11 13 15 17 19 3 5 7 9
389 > ls /proc/irq/0/
390 smp_affinity
392 The contents of the prof_cpu_mask file and each smp_affinity file for each IRQ
393 is the same by default:
395 > cat /proc/irq/0/smp_affinity
396 ffffffff
398 It's a bitmask, in which you can specify which CPUs can handle the IRQ, you can
399 set it by doing:
401 > echo 1 > /proc/irq/prof_cpu_mask
403 This means that only the first CPU will handle the IRQ, but you can also echo 5
404 which means that only the first and fourth CPU can handle the IRQ.
406 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
407 between all the CPUs which are allowed to handle it. As usual the kernel has
408 more info than you and does a better job than you, so the defaults are the
409 best choice for almost everyone.
411 There are three more important subdirectories in /proc: net, scsi, and sys.
412 The general rule is that the contents, or even the existence of these
413 directories, depend on your kernel configuration. If SCSI is not enabled, the
414 directory scsi may not exist. The same is true with the net, which is there
415 only when networking support is present in the running kernel.
417 The slabinfo file gives information about memory usage at the slab level.
418 Linux uses slab pools for memory management above page level in version 2.2.
419 Commonly used objects have their own slab pool (such as network buffers,
420 directory cache, and so on).
422 ..............................................................................
424 > cat /proc/buddyinfo
426 Node 0, zone DMA 0 4 5 4 4 3 ...
427 Node 0, zone Normal 1 0 0 1 101 8 ...
428 Node 0, zone HighMem 2 0 0 1 1 0 ...
430 Memory fragmentation is a problem under some workloads, and buddyinfo is a
431 useful tool for helping diagnose these problems. Buddyinfo will give you a
432 clue as to how big an area you can safely allocate, or why a previous
433 allocation failed.
435 Each column represents the number of pages of a certain order which are
436 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
437 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
438 available in ZONE_NORMAL, etc...
440 ..............................................................................
442 meminfo:
444 Provides information about distribution and utilization of memory. This
445 varies by architecture and compile options. The following is from a
446 16GB PIII, which has highmem enabled. You may not have all of these fields.
448 > cat /proc/meminfo
451 MemTotal: 16344972 kB
452 MemFree: 13634064 kB
453 Buffers: 3656 kB
454 Cached: 1195708 kB
455 SwapCached: 0 kB
456 Active: 891636 kB
457 Inactive: 1077224 kB
458 HighTotal: 15597528 kB
459 HighFree: 13629632 kB
460 LowTotal: 747444 kB
461 LowFree: 4432 kB
462 SwapTotal: 0 kB
463 SwapFree: 0 kB
464 Dirty: 968 kB
465 Writeback: 0 kB
466 AnonPages: 861800 kB
467 Mapped: 280372 kB
468 Slab: 284364 kB
469 SReclaimable: 159856 kB
470 SUnreclaim: 124508 kB
471 PageTables: 24448 kB
472 NFS_Unstable: 0 kB
473 Bounce: 0 kB
474 WritebackTmp: 0 kB
475 CommitLimit: 7669796 kB
476 Committed_AS: 100056 kB
477 VmallocTotal: 112216 kB
478 VmallocUsed: 428 kB
479 VmallocChunk: 111088 kB
481 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
482 bits and the kernel binary code)
483 MemFree: The sum of LowFree+HighFree
484 Buffers: Relatively temporary storage for raw disk blocks
485 shouldn't get tremendously large (20MB or so)
486 Cached: in-memory cache for files read from the disk (the
487 pagecache). Doesn't include SwapCached
488 SwapCached: Memory that once was swapped out, is swapped back in but
489 still also is in the swapfile (if memory is needed it
490 doesn't need to be swapped out AGAIN because it is already
491 in the swapfile. This saves I/O)
492 Active: Memory that has been used more recently and usually not
493 reclaimed unless absolutely necessary.
494 Inactive: Memory which has been less recently used. It is more
495 eligible to be reclaimed for other purposes
496 HighTotal:
497 HighFree: Highmem is all memory above ~860MB of physical memory
498 Highmem areas are for use by userspace programs, or
499 for the pagecache. The kernel must use tricks to access
500 this memory, making it slower to access than lowmem.
501 LowTotal:
502 LowFree: Lowmem is memory which can be used for everything that
503 highmem can be used for, but it is also available for the
504 kernel's use for its own data structures. Among many
505 other things, it is where everything from the Slab is
506 allocated. Bad things happen when you're out of lowmem.
507 SwapTotal: total amount of swap space available
508 SwapFree: Memory which has been evicted from RAM, and is temporarily
509 on the disk
510 Dirty: Memory which is waiting to get written back to the disk
511 Writeback: Memory which is actively being written back to the disk
512 AnonPages: Non-file backed pages mapped into userspace page tables
513 Mapped: files which have been mmaped, such as libraries
514 Slab: in-kernel data structures cache
515 SReclaimable: Part of Slab, that might be reclaimed, such as caches
516 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
517 PageTables: amount of memory dedicated to the lowest level of page
518 tables.
519 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
520 storage
521 Bounce: Memory used for block device "bounce buffers"
522 WritebackTmp: Memory used by FUSE for temporary writeback buffers
523 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
524 this is the total amount of memory currently available to
525 be allocated on the system. This limit is only adhered to
526 if strict overcommit accounting is enabled (mode 2 in
527 'vm.overcommit_memory').
528 The CommitLimit is calculated with the following formula:
529 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
530 For example, on a system with 1G of physical RAM and 7G
531 of swap with a `vm.overcommit_ratio` of 30 it would
532 yield a CommitLimit of 7.3G.
533 For more details, see the memory overcommit documentation
534 in vm/overcommit-accounting.
535 Committed_AS: The amount of memory presently allocated on the system.
536 The committed memory is a sum of all of the memory which
537 has been allocated by processes, even if it has not been
538 "used" by them as of yet. A process which malloc()'s 1G
539 of memory, but only touches 300M of it will only show up
540 as using 300M of memory even if it has the address space
541 allocated for the entire 1G. This 1G is memory which has
542 been "committed" to by the VM and can be used at any time
543 by the allocating application. With strict overcommit
544 enabled on the system (mode 2 in 'vm.overcommit_memory'),
545 allocations which would exceed the CommitLimit (detailed
546 above) will not be permitted. This is useful if one needs
547 to guarantee that processes will not fail due to lack of
548 memory once that memory has been successfully allocated.
549 VmallocTotal: total size of vmalloc memory area
550 VmallocUsed: amount of vmalloc area which is used
551 VmallocChunk: largest contigious block of vmalloc area which is free
554 1.3 IDE devices in /proc/ide
555 ----------------------------
557 The subdirectory /proc/ide contains information about all IDE devices of which
558 the kernel is aware. There is one subdirectory for each IDE controller, the
559 file drivers and a link for each IDE device, pointing to the device directory
560 in the controller specific subtree.
562 The file drivers contains general information about the drivers used for the
563 IDE devices:
565 > cat /proc/ide/drivers
566 ide-cdrom version 4.53
567 ide-disk version 1.08
569 More detailed information can be found in the controller specific
570 subdirectories. These are named ide0, ide1 and so on. Each of these
571 directories contains the files shown in table 1-5.
574 Table 1-5: IDE controller info in /proc/ide/ide?
575 ..............................................................................
576 File Content
577 channel IDE channel (0 or 1)
578 config Configuration (only for PCI/IDE bridge)
579 mate Mate name
580 model Type/Chipset of IDE controller
581 ..............................................................................
583 Each device connected to a controller has a separate subdirectory in the
584 controllers directory. The files listed in table 1-6 are contained in these
585 directories.
588 Table 1-6: IDE device information
589 ..............................................................................
590 File Content
591 cache The cache
592 capacity Capacity of the medium (in 512Byte blocks)
593 driver driver and version
594 geometry physical and logical geometry
595 identify device identify block
596 media media type
597 model device identifier
598 settings device setup
599 smart_thresholds IDE disk management thresholds
600 smart_values IDE disk management values
601 ..............................................................................
603 The most interesting file is settings. This file contains a nice overview of
604 the drive parameters:
606 # cat /proc/ide/ide0/hda/settings
607 name value min max mode
608 ---- ----- --- --- ----
609 bios_cyl 526 0 65535 rw
610 bios_head 255 0 255 rw
611 bios_sect 63 0 63 rw
612 breada_readahead 4 0 127 rw
613 bswap 0 0 1 r
614 file_readahead 72 0 2097151 rw
615 io_32bit 0 0 3 rw
616 keepsettings 0 0 1 rw
617 max_kb_per_request 122 1 127 rw
618 multcount 0 0 8 rw
619 nice1 1 0 1 rw
620 nowerr 0 0 1 rw
621 pio_mode write-only 0 255 w
622 slow 0 0 1 rw
623 unmaskirq 0 0 1 rw
624 using_dma 0 0 1 rw
627 1.4 Networking info in /proc/net
628 --------------------------------
630 The subdirectory /proc/net follows the usual pattern. Table 1-6 shows the
631 additional values you get for IP version 6 if you configure the kernel to
632 support this. Table 1-7 lists the files and their meaning.
635 Table 1-6: IPv6 info in /proc/net
636 ..............................................................................
637 File Content
638 udp6 UDP sockets (IPv6)
639 tcp6 TCP sockets (IPv6)
640 raw6 Raw device statistics (IPv6)
641 igmp6 IP multicast addresses, which this host joined (IPv6)
642 if_inet6 List of IPv6 interface addresses
643 ipv6_route Kernel routing table for IPv6
644 rt6_stats Global IPv6 routing tables statistics
645 sockstat6 Socket statistics (IPv6)
646 snmp6 Snmp data (IPv6)
647 ..............................................................................
650 Table 1-7: Network info in /proc/net
651 ..............................................................................
652 File Content
653 arp Kernel ARP table
654 dev network devices with statistics
655 dev_mcast the Layer2 multicast groups a device is listening too
656 (interface index, label, number of references, number of bound
657 addresses).
658 dev_stat network device status
659 ip_fwchains Firewall chain linkage
660 ip_fwnames Firewall chain names
661 ip_masq Directory containing the masquerading tables
662 ip_masquerade Major masquerading table
663 netstat Network statistics
664 raw raw device statistics
665 route Kernel routing table
666 rpc Directory containing rpc info
667 rt_cache Routing cache
668 snmp SNMP data
669 sockstat Socket statistics
670 tcp TCP sockets
671 tr_rif Token ring RIF routing table
672 udp UDP sockets
673 unix UNIX domain sockets
674 wireless Wireless interface data (Wavelan etc)
675 igmp IP multicast addresses, which this host joined
676 psched Global packet scheduler parameters.
677 netlink List of PF_NETLINK sockets
678 ip_mr_vifs List of multicast virtual interfaces
679 ip_mr_cache List of multicast routing cache
680 ..............................................................................
682 You can use this information to see which network devices are available in
683 your system and how much traffic was routed over those devices:
685 > cat /proc/net/dev
686 Inter-|Receive |[...
687 face |bytes packets errs drop fifo frame compressed multicast|[...
688 lo: 908188 5596 0 0 0 0 0 0 [...
689 ppp0:15475140 20721 410 0 0 410 0 0 [...
690 eth0: 614530 7085 0 0 0 0 0 1 [...
692 ...] Transmit
693 ...] bytes packets errs drop fifo colls carrier compressed
694 ...] 908188 5596 0 0 0 0 0 0
695 ...] 1375103 17405 0 0 0 0 0 0
696 ...] 1703981 5535 0 0 0 3 0 0
698 In addition, each Channel Bond interface has it's own directory. For
699 example, the bond0 device will have a directory called /proc/net/bond0/.
700 It will contain information that is specific to that bond, such as the
701 current slaves of the bond, the link status of the slaves, and how
702 many times the slaves link has failed.
704 1.5 SCSI info
705 -------------
707 If you have a SCSI host adapter in your system, you'll find a subdirectory
708 named after the driver for this adapter in /proc/scsi. You'll also see a list
709 of all recognized SCSI devices in /proc/scsi:
711 >cat /proc/scsi/scsi
712 Attached devices:
713 Host: scsi0 Channel: 00 Id: 00 Lun: 00
714 Vendor: IBM Model: DGHS09U Rev: 03E0
715 Type: Direct-Access ANSI SCSI revision: 03
716 Host: scsi0 Channel: 00 Id: 06 Lun: 00
717 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
718 Type: CD-ROM ANSI SCSI revision: 02
721 The directory named after the driver has one file for each adapter found in
722 the system. These files contain information about the controller, including
723 the used IRQ and the IO address range. The amount of information shown is
724 dependent on the adapter you use. The example shows the output for an Adaptec
725 AHA-2940 SCSI adapter:
727 > cat /proc/scsi/aic7xxx/0
729 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
730 Compile Options:
731 TCQ Enabled By Default : Disabled
732 AIC7XXX_PROC_STATS : Disabled
733 AIC7XXX_RESET_DELAY : 5
734 Adapter Configuration:
735 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
736 Ultra Wide Controller
737 PCI MMAPed I/O Base: 0xeb001000
738 Adapter SEEPROM Config: SEEPROM found and used.
739 Adaptec SCSI BIOS: Enabled
740 IRQ: 10
741 SCBs: Active 0, Max Active 2,
742 Allocated 15, HW 16, Page 255
743 Interrupts: 160328
744 BIOS Control Word: 0x18b6
745 Adapter Control Word: 0x005b
746 Extended Translation: Enabled
747 Disconnect Enable Flags: 0xffff
748 Ultra Enable Flags: 0x0001
749 Tag Queue Enable Flags: 0x0000
750 Ordered Queue Tag Flags: 0x0000
751 Default Tag Queue Depth: 8
752 Tagged Queue By Device array for aic7xxx host instance 0:
753 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
754 Actual queue depth per device for aic7xxx host instance 0:
755 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
756 Statistics:
757 (scsi0:0:0:0)
758 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
759 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
760 Total transfers 160151 (74577 reads and 85574 writes)
761 (scsi0:0:6:0)
762 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
763 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
764 Total transfers 0 (0 reads and 0 writes)
767 1.6 Parallel port info in /proc/parport
768 ---------------------------------------
770 The directory /proc/parport contains information about the parallel ports of
771 your system. It has one subdirectory for each port, named after the port
772 number (0,1,2,...).
774 These directories contain the four files shown in Table 1-8.
777 Table 1-8: Files in /proc/parport
778 ..............................................................................
779 File Content
780 autoprobe Any IEEE-1284 device ID information that has been acquired.
781 devices list of the device drivers using that port. A + will appear by the
782 name of the device currently using the port (it might not appear
783 against any).
784 hardware Parallel port's base address, IRQ line and DMA channel.
785 irq IRQ that parport is using for that port. This is in a separate
786 file to allow you to alter it by writing a new value in (IRQ
787 number or none).
788 ..............................................................................
790 1.7 TTY info in /proc/tty
791 -------------------------
793 Information about the available and actually used tty's can be found in the
794 directory /proc/tty.You'll find entries for drivers and line disciplines in
795 this directory, as shown in Table 1-9.
798 Table 1-9: Files in /proc/tty
799 ..............................................................................
800 File Content
801 drivers list of drivers and their usage
802 ldiscs registered line disciplines
803 driver/serial usage statistic and status of single tty lines
804 ..............................................................................
806 To see which tty's are currently in use, you can simply look into the file
807 /proc/tty/drivers:
809 > cat /proc/tty/drivers
810 pty_slave /dev/pts 136 0-255 pty:slave
811 pty_master /dev/ptm 128 0-255 pty:master
812 pty_slave /dev/ttyp 3 0-255 pty:slave
813 pty_master /dev/pty 2 0-255 pty:master
814 serial /dev/cua 5 64-67 serial:callout
815 serial /dev/ttyS 4 64-67 serial
816 /dev/tty0 /dev/tty0 4 0 system:vtmaster
817 /dev/ptmx /dev/ptmx 5 2 system
818 /dev/console /dev/console 5 1 system:console
819 /dev/tty /dev/tty 5 0 system:/dev/tty
820 unknown /dev/tty 4 1-63 console
823 1.8 Miscellaneous kernel statistics in /proc/stat
824 -------------------------------------------------
826 Various pieces of information about kernel activity are available in the
827 /proc/stat file. All of the numbers reported in this file are aggregates
828 since the system first booted. For a quick look, simply cat the file:
830 > cat /proc/stat
831 cpu 2255 34 2290 22625563 6290 127 456 0
832 cpu0 1132 34 1441 11311718 3675 127 438 0
833 cpu1 1123 0 849 11313845 2614 0 18 0
834 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
835 ctxt 1990473
836 btime 1062191376
837 processes 2915
838 procs_running 1
839 procs_blocked 0
841 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
842 lines. These numbers identify the amount of time the CPU has spent performing
843 different kinds of work. Time units are in USER_HZ (typically hundredths of a
844 second). The meanings of the columns are as follows, from left to right:
846 - user: normal processes executing in user mode
847 - nice: niced processes executing in user mode
848 - system: processes executing in kernel mode
849 - idle: twiddling thumbs
850 - iowait: waiting for I/O to complete
851 - irq: servicing interrupts
852 - softirq: servicing softirqs
853 - steal: involuntary wait
855 The "intr" line gives counts of interrupts serviced since boot time, for each
856 of the possible system interrupts. The first column is the total of all
857 interrupts serviced; each subsequent column is the total for that particular
858 interrupt.
860 The "ctxt" line gives the total number of context switches across all CPUs.
862 The "btime" line gives the time at which the system booted, in seconds since
863 the Unix epoch.
865 The "processes" line gives the number of processes and threads created, which
866 includes (but is not limited to) those created by calls to the fork() and
867 clone() system calls.
869 The "procs_running" line gives the number of processes currently running on
870 CPUs.
872 The "procs_blocked" line gives the number of processes currently blocked,
873 waiting for I/O to complete.
875 1.9 Ext4 file system parameters
876 ------------------------------
877 Ext4 file system have one directory per partition under /proc/fs/ext4/
878 # ls /proc/fs/ext4/hdc/
879 group_prealloc max_to_scan mb_groups mb_history min_to_scan order2_req
880 stats stream_req
882 mb_groups:
883 This file gives the details of mutiblock allocator buddy cache of free blocks
885 mb_history:
886 Multiblock allocation history.
888 stats:
889 This file indicate whether the multiblock allocator should start collecting
890 statistics. The statistics are shown during unmount
892 group_prealloc:
893 The multiblock allocator normalize the block allocation request to
894 group_prealloc filesystem blocks if we don't have strip value set.
895 The stripe value can be specified at mount time or during mke2fs.
897 max_to_scan:
898 How long multiblock allocator can look for a best extent (in found extents)
900 min_to_scan:
901 How long multiblock allocator must look for a best extent
903 order2_req:
904 Multiblock allocator use 2^N search using buddies only for requests greater
905 than or equal to order2_req. The request size is specfied in file system
906 blocks. A value of 2 indicate only if the requests are greater than or equal
907 to 4 blocks.
909 stream_req:
910 Files smaller than stream_req are served by the stream allocator, whose
911 purpose is to pack requests as close each to other as possible to
912 produce smooth I/O traffic. Avalue of 16 indicate that file smaller than 16
913 filesystem block size will use group based preallocation.
915 ------------------------------------------------------------------------------
916 Summary
917 ------------------------------------------------------------------------------
918 The /proc file system serves information about the running system. It not only
919 allows access to process data but also allows you to request the kernel status
920 by reading files in the hierarchy.
922 The directory structure of /proc reflects the types of information and makes
923 it easy, if not obvious, where to look for specific data.
924 ------------------------------------------------------------------------------
926 ------------------------------------------------------------------------------
927 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
928 ------------------------------------------------------------------------------
930 ------------------------------------------------------------------------------
931 In This Chapter
932 ------------------------------------------------------------------------------
933 * Modifying kernel parameters by writing into files found in /proc/sys
934 * Exploring the files which modify certain parameters
935 * Review of the /proc/sys file tree
936 ------------------------------------------------------------------------------
939 A very interesting part of /proc is the directory /proc/sys. This is not only
940 a source of information, it also allows you to change parameters within the
941 kernel. Be very careful when attempting this. You can optimize your system,
942 but you can also cause it to crash. Never alter kernel parameters on a
943 production system. Set up a development machine and test to make sure that
944 everything works the way you want it to. You may have no alternative but to
945 reboot the machine once an error has been made.
947 To change a value, simply echo the new value into the file. An example is
948 given below in the section on the file system data. You need to be root to do
949 this. You can create your own boot script to perform this every time your
950 system boots.
952 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
953 general things in the operation of the Linux kernel. Since some of the files
954 can inadvertently disrupt your system, it is advisable to read both
955 documentation and source before actually making adjustments. In any case, be
956 very careful when writing to any of these files. The entries in /proc may
957 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
958 review the kernel documentation in the directory /usr/src/linux/Documentation.
959 This chapter is heavily based on the documentation included in the pre 2.2
960 kernels, and became part of it in version 2.2.1 of the Linux kernel.
962 2.1 /proc/sys/fs - File system data
963 -----------------------------------
965 This subdirectory contains specific file system, file handle, inode, dentry
966 and quota information.
968 Currently, these files are in /proc/sys/fs:
970 dentry-state
971 ------------
973 Status of the directory cache. Since directory entries are dynamically
974 allocated and deallocated, this file indicates the current status. It holds
975 six values, in which the last two are not used and are always zero. The others
976 are listed in table 2-1.
979 Table 2-1: Status files of the directory cache
980 ..............................................................................
981 File Content
982 nr_dentry Almost always zero
983 nr_unused Number of unused cache entries
984 age_limit
985 in seconds after the entry may be reclaimed, when memory is short
986 want_pages internally
987 ..............................................................................
989 dquot-nr and dquot-max
990 ----------------------
992 The file dquot-max shows the maximum number of cached disk quota entries.
994 The file dquot-nr shows the number of allocated disk quota entries and the
995 number of free disk quota entries.
997 If the number of available cached disk quotas is very low and you have a large
998 number of simultaneous system users, you might want to raise the limit.
1000 file-nr and file-max
1001 --------------------
1003 The kernel allocates file handles dynamically, but doesn't free them again at
1004 this time.
1006 The value in file-max denotes the maximum number of file handles that the
1007 Linux kernel will allocate. When you get a lot of error messages about running
1008 out of file handles, you might want to raise this limit. The default value is
1009 10% of RAM in kilobytes. To change it, just write the new number into the
1010 file:
1012 # cat /proc/sys/fs/file-max
1013 4096
1014 # echo 8192 > /proc/sys/fs/file-max
1015 # cat /proc/sys/fs/file-max
1016 8192
1019 This method of revision is useful for all customizable parameters of the
1020 kernel - simply echo the new value to the corresponding file.
1022 Historically, the three values in file-nr denoted the number of allocated file
1023 handles, the number of allocated but unused file handles, and the maximum
1024 number of file handles. Linux 2.6 always reports 0 as the number of free file
1025 handles -- this is not an error, it just means that the number of allocated
1026 file handles exactly matches the number of used file handles.
1028 Attempts to allocate more file descriptors than file-max are reported with
1029 printk, look for "VFS: file-max limit <number> reached".
1031 inode-state and inode-nr
1032 ------------------------
1034 The file inode-nr contains the first two items from inode-state, so we'll skip
1035 to that file...
1037 inode-state contains two actual numbers and five dummy values. The numbers
1038 are nr_inodes and nr_free_inodes (in order of appearance).
1040 nr_inodes
1041 ~~~~~~~~~
1043 Denotes the number of inodes the system has allocated. This number will
1044 grow and shrink dynamically.
1046 nr_open
1047 -------
1049 Denotes the maximum number of file-handles a process can
1050 allocate. Default value is 1024*1024 (1048576) which should be
1051 enough for most machines. Actual limit depends on RLIMIT_NOFILE
1052 resource limit.
1054 nr_free_inodes
1055 --------------
1057 Represents the number of free inodes. Ie. The number of inuse inodes is
1058 (nr_inodes - nr_free_inodes).
1060 aio-nr and aio-max-nr
1061 ---------------------
1063 aio-nr is the running total of the number of events specified on the
1064 io_setup system call for all currently active aio contexts. If aio-nr
1065 reaches aio-max-nr then io_setup will fail with EAGAIN. Note that
1066 raising aio-max-nr does not result in the pre-allocation or re-sizing
1067 of any kernel data structures.
1069 2.2 /proc/sys/fs/binfmt_misc - Miscellaneous binary formats
1070 -----------------------------------------------------------
1072 Besides these files, there is the subdirectory /proc/sys/fs/binfmt_misc. This
1073 handles the kernel support for miscellaneous binary formats.
1075 Binfmt_misc provides the ability to register additional binary formats to the
1076 Kernel without compiling an additional module/kernel. Therefore, binfmt_misc
1077 needs to know magic numbers at the beginning or the filename extension of the
1078 binary.
1080 It works by maintaining a linked list of structs that contain a description of
1081 a binary format, including a magic with size (or the filename extension),
1082 offset and mask, and the interpreter name. On request it invokes the given
1083 interpreter with the original program as argument, as binfmt_java and
1084 binfmt_em86 and binfmt_mz do. Since binfmt_misc does not define any default
1085 binary-formats, you have to register an additional binary-format.
1087 There are two general files in binfmt_misc and one file per registered format.
1088 The two general files are register and status.
1090 Registering a new binary format
1091 -------------------------------
1093 To register a new binary format you have to issue the command
1095 echo :name:type:offset:magic:mask:interpreter: > /proc/sys/fs/binfmt_misc/register
1099 with appropriate name (the name for the /proc-dir entry), offset (defaults to
1100 0, if omitted), magic, mask (which can be omitted, defaults to all 0xff) and
1101 last but not least, the interpreter that is to be invoked (for example and
1102 testing /bin/echo). Type can be M for usual magic matching or E for filename
1103 extension matching (give extension in place of magic).
1105 Check or reset the status of the binary format handler
1106 ------------------------------------------------------
1108 If you do a cat on the file /proc/sys/fs/binfmt_misc/status, you will get the
1109 current status (enabled/disabled) of binfmt_misc. Change the status by echoing
1110 0 (disables) or 1 (enables) or -1 (caution: this clears all previously
1111 registered binary formats) to status. For example echo 0 > status to disable
1112 binfmt_misc (temporarily).
1114 Status of a single handler
1115 --------------------------
1117 Each registered handler has an entry in /proc/sys/fs/binfmt_misc. These files
1118 perform the same function as status, but their scope is limited to the actual
1119 binary format. By cating this file, you also receive all related information
1120 about the interpreter/magic of the binfmt.
1122 Example usage of binfmt_misc (emulate binfmt_java)
1123 --------------------------------------------------
1125 cd /proc/sys/fs/binfmt_misc
1126 echo ':Java:M::\xca\xfe\xba\xbe::/usr/local/java/bin/javawrapper:' > register
1127 echo ':HTML:E::html::/usr/local/java/bin/appletviewer:' > register
1128 echo ':Applet:M::<!--applet::/usr/local/java/bin/appletviewer:' > register
1129 echo ':DEXE:M::\x0eDEX::/usr/bin/dosexec:' > register
1132 These four lines add support for Java executables and Java applets (like
1133 binfmt_java, additionally recognizing the .html extension with no need to put
1134 <!--applet> to every applet file). You have to install the JDK and the
1135 shell-script /usr/local/java/bin/javawrapper too. It works around the
1136 brokenness of the Java filename handling. To add a Java binary, just create a
1137 link to the class-file somewhere in the path.
1139 2.3 /proc/sys/kernel - general kernel parameters
1140 ------------------------------------------------
1142 This directory reflects general kernel behaviors. As I've said before, the
1143 contents depend on your configuration. Here you'll find the most important
1144 files, along with descriptions of what they mean and how to use them.
1146 acct
1147 ----
1149 The file contains three values; highwater, lowwater, and frequency.
1151 It exists only when BSD-style process accounting is enabled. These values
1152 control its behavior. If the free space on the file system where the log lives
1153 goes below lowwater percentage, accounting suspends. If it goes above
1154 highwater percentage, accounting resumes. Frequency determines how often you
1155 check the amount of free space (value is in seconds). Default settings are: 4,
1156 2, and 30. That is, suspend accounting if there is less than 2 percent free;
1157 resume it if we have a value of 3 or more percent; consider information about
1158 the amount of free space valid for 30 seconds
1160 ctrl-alt-del
1161 ------------
1163 When the value in this file is 0, ctrl-alt-del is trapped and sent to the init
1164 program to handle a graceful restart. However, when the value is greater that
1165 zero, Linux's reaction to this key combination will be an immediate reboot,
1166 without syncing its dirty buffers.
1168 [NOTE]
1169 When a program (like dosemu) has the keyboard in raw mode, the
1170 ctrl-alt-del is intercepted by the program before it ever reaches the
1171 kernel tty layer, and it is up to the program to decide what to do with
1172 it.
1174 domainname and hostname
1175 -----------------------
1177 These files can be controlled to set the NIS domainname and hostname of your
1178 box. For the classic darkstar.frop.org a simple:
1180 # echo "darkstar" > /proc/sys/kernel/hostname
1181 # echo "frop.org" > /proc/sys/kernel/domainname
1184 would suffice to set your hostname and NIS domainname.
1186 osrelease, ostype and version
1187 -----------------------------
1189 The names make it pretty obvious what these fields contain:
1191 > cat /proc/sys/kernel/osrelease
1192 2.2.12
1194 > cat /proc/sys/kernel/ostype
1195 Linux
1197 > cat /proc/sys/kernel/version
1198 #4 Fri Oct 1 12:41:14 PDT 1999
1201 The files osrelease and ostype should be clear enough. Version needs a little
1202 more clarification. The #4 means that this is the 4th kernel built from this
1203 source base and the date after it indicates the time the kernel was built. The
1204 only way to tune these values is to rebuild the kernel.
1206 panic
1207 -----
1209 The value in this file represents the number of seconds the kernel waits
1210 before rebooting on a panic. When you use the software watchdog, the
1211 recommended setting is 60. If set to 0, the auto reboot after a kernel panic
1212 is disabled, which is the default setting.
1214 printk
1215 ------
1217 The four values in printk denote
1218 * console_loglevel,
1219 * default_message_loglevel,
1220 * minimum_console_loglevel and
1221 * default_console_loglevel
1222 respectively.
1224 These values influence printk() behavior when printing or logging error
1225 messages, which come from inside the kernel. See syslog(2) for more
1226 information on the different log levels.
1228 console_loglevel
1229 ----------------
1231 Messages with a higher priority than this will be printed to the console.
1233 default_message_level
1234 ---------------------
1236 Messages without an explicit priority will be printed with this priority.
1238 minimum_console_loglevel
1239 ------------------------
1241 Minimum (highest) value to which the console_loglevel can be set.
1243 default_console_loglevel
1244 ------------------------
1246 Default value for console_loglevel.
1248 sg-big-buff
1249 -----------
1251 This file shows the size of the generic SCSI (sg) buffer. At this point, you
1252 can't tune it yet, but you can change it at compile time by editing
1253 include/scsi/sg.h and changing the value of SG_BIG_BUFF.
1255 If you use a scanner with SANE (Scanner Access Now Easy) you might want to set
1256 this to a higher value. Refer to the SANE documentation on this issue.
1258 modprobe
1259 --------
1261 The location where the modprobe binary is located. The kernel uses this
1262 program to load modules on demand.
1264 unknown_nmi_panic
1265 -----------------
1267 The value in this file affects behavior of handling NMI. When the value is
1268 non-zero, unknown NMI is trapped and then panic occurs. At that time, kernel
1269 debugging information is displayed on console.
1271 NMI switch that most IA32 servers have fires unknown NMI up, for example.
1272 If a system hangs up, try pressing the NMI switch.
1274 nmi_watchdog
1275 ------------
1277 Enables/Disables the NMI watchdog on x86 systems. When the value is non-zero
1278 the NMI watchdog is enabled and will continuously test all online cpus to
1279 determine whether or not they are still functioning properly.
1281 Because the NMI watchdog shares registers with oprofile, by disabling the NMI
1282 watchdog, oprofile may have more registers to utilize.
1284 maps_protect
1285 ------------
1287 Enables/Disables the protection of the per-process proc entries "maps" and
1288 "smaps". When enabled, the contents of these files are visible only to
1289 readers that are allowed to ptrace() the given process.
1292 2.4 /proc/sys/vm - The virtual memory subsystem
1293 -----------------------------------------------
1295 The files in this directory can be used to tune the operation of the virtual
1296 memory (VM) subsystem of the Linux kernel.
1298 vfs_cache_pressure
1299 ------------------
1301 Controls the tendency of the kernel to reclaim the memory which is used for
1302 caching of directory and inode objects.
1304 At the default value of vfs_cache_pressure=100 the kernel will attempt to
1305 reclaim dentries and inodes at a "fair" rate with respect to pagecache and
1306 swapcache reclaim. Decreasing vfs_cache_pressure causes the kernel to prefer
1307 to retain dentry and inode caches. Increasing vfs_cache_pressure beyond 100
1308 causes the kernel to prefer to reclaim dentries and inodes.
1310 dirty_background_ratio
1311 ----------------------
1313 Contains, as a percentage of total system memory, the number of pages at which
1314 the pdflush background writeback daemon will start writing out dirty data.
1316 dirty_ratio
1317 -----------------
1319 Contains, as a percentage of total system memory, the number of pages at which
1320 a process which is generating disk writes will itself start writing out dirty
1321 data.
1323 dirty_writeback_centisecs
1324 -------------------------
1326 The pdflush writeback daemons will periodically wake up and write `old' data
1327 out to disk. This tunable expresses the interval between those wakeups, in
1328 100'ths of a second.
1330 Setting this to zero disables periodic writeback altogether.
1332 dirty_expire_centisecs
1333 ----------------------
1335 This tunable is used to define when dirty data is old enough to be eligible
1336 for writeout by the pdflush daemons. It is expressed in 100'ths of a second.
1337 Data which has been dirty in-memory for longer than this interval will be
1338 written out next time a pdflush daemon wakes up.
1340 highmem_is_dirtyable
1341 --------------------
1343 Only present if CONFIG_HIGHMEM is set.
1345 This defaults to 0 (false), meaning that the ratios set above are calculated
1346 as a percentage of lowmem only. This protects against excessive scanning
1347 in page reclaim, swapping and general VM distress.
1349 Setting this to 1 can be useful on 32 bit machines where you want to make
1350 random changes within an MMAPed file that is larger than your available
1351 lowmem without causing large quantities of random IO. Is is safe if the
1352 behavior of all programs running on the machine is known and memory will
1353 not be otherwise stressed.
1355 legacy_va_layout
1356 ----------------
1358 If non-zero, this sysctl disables the new 32-bit mmap mmap layout - the kernel
1359 will use the legacy (2.4) layout for all processes.
1361 lowmem_reserve_ratio
1362 ---------------------
1364 For some specialised workloads on highmem machines it is dangerous for
1365 the kernel to allow process memory to be allocated from the "lowmem"
1366 zone. This is because that memory could then be pinned via the mlock()
1367 system call, or by unavailability of swapspace.
1369 And on large highmem machines this lack of reclaimable lowmem memory
1370 can be fatal.
1372 So the Linux page allocator has a mechanism which prevents allocations
1373 which _could_ use highmem from using too much lowmem. This means that
1374 a certain amount of lowmem is defended from the possibility of being
1375 captured into pinned user memory.
1377 (The same argument applies to the old 16 megabyte ISA DMA region. This
1378 mechanism will also defend that region from allocations which could use
1379 highmem or lowmem).
1381 The `lowmem_reserve_ratio' tunable determines how aggressive the kernel is
1382 in defending these lower zones.
1384 If you have a machine which uses highmem or ISA DMA and your
1385 applications are using mlock(), or if you are running with no swap then
1386 you probably should change the lowmem_reserve_ratio setting.
1388 The lowmem_reserve_ratio is an array. You can see them by reading this file.
1389 -
1390 % cat /proc/sys/vm/lowmem_reserve_ratio
1391 256 256 32
1392 -
1393 Note: # of this elements is one fewer than number of zones. Because the highest
1394 zone's value is not necessary for following calculation.
1396 But, these values are not used directly. The kernel calculates # of protection
1397 pages for each zones from them. These are shown as array of protection pages
1398 in /proc/zoneinfo like followings. (This is an example of x86-64 box).
1399 Each zone has an array of protection pages like this.
1401 -
1402 Node 0, zone DMA
1403 pages free 1355
1404 min 3
1405 low 3
1406 high 4
1407 :
1408 :
1409 numa_other 0
1410 protection: (0, 2004, 2004, 2004)
1411 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1412 pagesets
1413 cpu: 0 pcp: 0
1414 :
1415 -
1416 These protections are added to score to judge whether this zone should be used
1417 for page allocation or should be reclaimed.
1419 In this example, if normal pages (index=2) are required to this DMA zone and
1420 pages_high is used for watermark, the kernel judges this zone should not be
1421 used because pages_free(1355) is smaller than watermark + protection[2]
1422 (4 + 2004 = 2008). If this protection value is 0, this zone would be used for
1423 normal page requirement. If requirement is DMA zone(index=0), protection[0]
1424 (=0) is used.
1426 zone[i]'s protection[j] is calculated by following exprssion.
1428 (i < j):
1429 zone[i]->protection[j]
1430 = (total sums of present_pages from zone[i+1] to zone[j] on the node)
1431 / lowmem_reserve_ratio[i];
1432 (i = j):
1433 (should not be protected. = 0;
1434 (i > j):
1435 (not necessary, but looks 0)
1437 The default values of lowmem_reserve_ratio[i] are
1438 256 (if zone[i] means DMA or DMA32 zone)
1439 32 (others).
1440 As above expression, they are reciprocal number of ratio.
1441 256 means 1/256. # of protection pages becomes about "0.39%" of total present
1442 pages of higher zones on the node.
1444 If you would like to protect more pages, smaller values are effective.
1445 The minimum value is 1 (1/1 -> 100%).
1447 page-cluster
1448 ------------
1450 page-cluster controls the number of pages which are written to swap in
1451 a single attempt. The swap I/O size.
1453 It is a logarithmic value - setting it to zero means "1 page", setting
1454 it to 1 means "2 pages", setting it to 2 means "4 pages", etc.
1456 The default value is three (eight pages at a time). There may be some
1457 small benefits in tuning this to a different value if your workload is
1458 swap-intensive.
1460 overcommit_memory
1461 -----------------
1463 Controls overcommit of system memory, possibly allowing processes
1464 to allocate (but not use) more memory than is actually available.
1467 0 - Heuristic overcommit handling. Obvious overcommits of
1468 address space are refused. Used for a typical system. It
1469 ensures a seriously wild allocation fails while allowing
1470 overcommit to reduce swap usage. root is allowed to
1471 allocate slightly more memory in this mode. This is the
1472 default.
1474 1 - Always overcommit. Appropriate for some scientific
1475 applications.
1477 2 - Don't overcommit. The total address space commit
1478 for the system is not permitted to exceed swap plus a
1479 configurable percentage (default is 50) of physical RAM.
1480 Depending on the percentage you use, in most situations
1481 this means a process will not be killed while attempting
1482 to use already-allocated memory but will receive errors
1483 on memory allocation as appropriate.
1485 overcommit_ratio
1486 ----------------
1488 Percentage of physical memory size to include in overcommit calculations
1489 (see above.)
1491 Memory allocation limit = swapspace + physmem * (overcommit_ratio / 100)
1493 swapspace = total size of all swap areas
1494 physmem = size of physical memory in system
1496 nr_hugepages and hugetlb_shm_group
1497 ----------------------------------
1499 nr_hugepages configures number of hugetlb page reserved for the system.
1501 hugetlb_shm_group contains group id that is allowed to create SysV shared
1502 memory segment using hugetlb page.
1504 hugepages_treat_as_movable
1505 --------------------------
1507 This parameter is only useful when kernelcore= is specified at boot time to
1508 create ZONE_MOVABLE for pages that may be reclaimed or migrated. Huge pages
1509 are not movable so are not normally allocated from ZONE_MOVABLE. A non-zero
1510 value written to hugepages_treat_as_movable allows huge pages to be allocated
1511 from ZONE_MOVABLE.
1513 Once enabled, the ZONE_MOVABLE is treated as an area of memory the huge
1514 pages pool can easily grow or shrink within. Assuming that applications are
1515 not running that mlock() a lot of memory, it is likely the huge pages pool
1516 can grow to the size of ZONE_MOVABLE by repeatedly entering the desired value
1517 into nr_hugepages and triggering page reclaim.
1519 laptop_mode
1520 -----------
1522 laptop_mode is a knob that controls "laptop mode". All the things that are
1523 controlled by this knob are discussed in Documentation/laptops/laptop-mode.txt.
1525 block_dump
1526 ----------
1528 block_dump enables block I/O debugging when set to a nonzero value. More
1529 information on block I/O debugging is in Documentation/laptops/laptop-mode.txt.
1531 swap_token_timeout
1532 ------------------
1534 This file contains valid hold time of swap out protection token. The Linux
1535 VM has token based thrashing control mechanism and uses the token to prevent
1536 unnecessary page faults in thrashing situation. The unit of the value is
1537 second. The value would be useful to tune thrashing behavior.
1539 drop_caches
1540 -----------
1542 Writing to this will cause the kernel to drop clean caches, dentries and
1543 inodes from memory, causing that memory to become free.
1545 To free pagecache:
1546 echo 1 > /proc/sys/vm/drop_caches
1547 To free dentries and inodes:
1548 echo 2 > /proc/sys/vm/drop_caches
1549 To free pagecache, dentries and inodes:
1550 echo 3 > /proc/sys/vm/drop_caches
1552 As this is a non-destructive operation and dirty objects are not freeable, the
1553 user should run `sync' first.
1556 2.5 /proc/sys/dev - Device specific parameters
1557 ----------------------------------------------
1559 Currently there is only support for CDROM drives, and for those, there is only
1560 one read-only file containing information about the CD-ROM drives attached to
1561 the system:
1563 >cat /proc/sys/dev/cdrom/info
1564 CD-ROM information, Id: cdrom.c 2.55 1999/04/25
1566 drive name: sr0 hdb
1567 drive speed: 32 40
1568 drive # of slots: 1 0
1569 Can close tray: 1 1
1570 Can open tray: 1 1
1571 Can lock tray: 1 1
1572 Can change speed: 1 1
1573 Can select disk: 0 1
1574 Can read multisession: 1 1
1575 Can read MCN: 1 1
1576 Reports media changed: 1 1
1577 Can play audio: 1 1
1580 You see two drives, sr0 and hdb, along with a list of their features.
1582 2.6 /proc/sys/sunrpc - Remote procedure calls
1583 ---------------------------------------------
1585 This directory contains four files, which enable or disable debugging for the
1586 RPC functions NFS, NFS-daemon, RPC and NLM. The default values are 0. They can
1587 be set to one to turn debugging on. (The default value is 0 for each)
1589 2.7 /proc/sys/net - Networking stuff
1590 ------------------------------------
1592 The interface to the networking parts of the kernel is located in
1593 /proc/sys/net. Table 2-3 shows all possible subdirectories. You may see only
1594 some of them, depending on your kernel's configuration.
1597 Table 2-3: Subdirectories in /proc/sys/net
1598 ..............................................................................
1599 Directory Content Directory Content
1600 core General parameter appletalk Appletalk protocol
1601 unix Unix domain sockets netrom NET/ROM
1602 802 E802 protocol ax25 AX25
1603 ethernet Ethernet protocol rose X.25 PLP layer
1604 ipv4 IP version 4 x25 X.25 protocol
1605 ipx IPX token-ring IBM token ring
1606 bridge Bridging decnet DEC net
1607 ipv6 IP version 6
1608 ..............................................................................
1610 We will concentrate on IP networking here. Since AX15, X.25, and DEC Net are
1611 only minor players in the Linux world, we'll skip them in this chapter. You'll
1612 find some short info on Appletalk and IPX further on in this chapter. Review
1613 the online documentation and the kernel source to get a detailed view of the
1614 parameters for those protocols. In this section we'll discuss the
1615 subdirectories printed in bold letters in the table above. As default values
1616 are suitable for most needs, there is no need to change these values.
1618 /proc/sys/net/core - Network core options
1619 -----------------------------------------
1621 rmem_default
1622 ------------
1624 The default setting of the socket receive buffer in bytes.
1626 rmem_max
1627 --------
1629 The maximum receive socket buffer size in bytes.
1631 wmem_default
1632 ------------
1634 The default setting (in bytes) of the socket send buffer.
1636 wmem_max
1637 --------
1639 The maximum send socket buffer size in bytes.
1641 message_burst and message_cost
1642 ------------------------------
1644 These parameters are used to limit the warning messages written to the kernel
1645 log from the networking code. They enforce a rate limit to make a
1646 denial-of-service attack impossible. A higher message_cost factor, results in
1647 fewer messages that will be written. Message_burst controls when messages will
1648 be dropped. The default settings limit warning messages to one every five
1649 seconds.
1651 warnings
1652 --------
1654 This controls console messages from the networking stack that can occur because
1655 of problems on the network like duplicate address or bad checksums. Normally,
1656 this should be enabled, but if the problem persists the messages can be
1657 disabled.
1660 netdev_max_backlog
1661 ------------------
1663 Maximum number of packets, queued on the INPUT side, when the interface
1664 receives packets faster than kernel can process them.
1666 optmem_max
1667 ----------
1669 Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
1670 of struct cmsghdr structures with appended data.
1672 /proc/sys/net/unix - Parameters for Unix domain sockets
1673 -------------------------------------------------------
1675 There are only two files in this subdirectory. They control the delays for
1676 deleting and destroying socket descriptors.
1678 2.8 /proc/sys/net/ipv4 - IPV4 settings
1679 --------------------------------------
1681 IP version 4 is still the most used protocol in Unix networking. It will be
1682 replaced by IP version 6 in the next couple of years, but for the moment it's
1683 the de facto standard for the internet and is used in most networking
1684 environments around the world. Because of the importance of this protocol,
1685 we'll have a deeper look into the subtree controlling the behavior of the IPv4
1686 subsystem of the Linux kernel.
1688 Let's start with the entries in /proc/sys/net/ipv4.
1690 ICMP settings
1691 -------------
1693 icmp_echo_ignore_all and icmp_echo_ignore_broadcasts
1694 ----------------------------------------------------
1696 Turn on (1) or off (0), if the kernel should ignore all ICMP ECHO requests, or
1697 just those to broadcast and multicast addresses.
1699 Please note that if you accept ICMP echo requests with a broadcast/multi\-cast
1700 destination address your network may be used as an exploder for denial of
1701 service packet flooding attacks to other hosts.
1703 icmp_destunreach_rate, icmp_echoreply_rate, icmp_paramprob_rate and icmp_timeexeed_rate
1704 ---------------------------------------------------------------------------------------
1706 Sets limits for sending ICMP packets to specific targets. A value of zero
1707 disables all limiting. Any positive value sets the maximum package rate in
1708 hundredth of a second (on Intel systems).
1710 IP settings
1711 -----------
1713 ip_autoconfig
1714 -------------
1716 This file contains the number one if the host received its IP configuration by
1717 RARP, BOOTP, DHCP or a similar mechanism. Otherwise it is zero.
1719 ip_default_ttl
1720 --------------
1722 TTL (Time To Live) for IPv4 interfaces. This is simply the maximum number of
1723 hops a packet may travel.
1725 ip_dynaddr
1726 ----------
1728 Enable dynamic socket address rewriting on interface address change. This is
1729 useful for dialup interface with changing IP addresses.
1731 ip_forward
1732 ----------
1734 Enable or disable forwarding of IP packages between interfaces. Changing this
1735 value resets all other parameters to their default values. They differ if the
1736 kernel is configured as host or router.
1738 ip_local_port_range
1739 -------------------
1741 Range of ports used by TCP and UDP to choose the local port. Contains two
1742 numbers, the first number is the lowest port, the second number the highest
1743 local port. Default is 1024-4999. Should be changed to 32768-61000 for
1744 high-usage systems.
1746 ip_no_pmtu_disc
1747 ---------------
1749 Global switch to turn path MTU discovery off. It can also be set on a per
1750 socket basis by the applications or on a per route basis.
1752 ip_masq_debug
1753 -------------
1755 Enable/disable debugging of IP masquerading.
1757 IP fragmentation settings
1758 -------------------------
1760 ipfrag_high_trash and ipfrag_low_trash
1761 --------------------------------------
1763 Maximum memory used to reassemble IP fragments. When ipfrag_high_thresh bytes
1764 of memory is allocated for this purpose, the fragment handler will toss
1765 packets until ipfrag_low_thresh is reached.
1767 ipfrag_time
1768 -----------
1770 Time in seconds to keep an IP fragment in memory.
1772 TCP settings
1773 ------------
1775 tcp_ecn
1776 -------
1778 This file controls the use of the ECN bit in the IPv4 headers. This is a new
1779 feature about Explicit Congestion Notification, but some routers and firewalls
1780 block traffic that has this bit set, so it could be necessary to echo 0 to
1781 /proc/sys/net/ipv4/tcp_ecn if you want to talk to these sites. For more info
1782 you could read RFC2481.
1784 tcp_retrans_collapse
1785 --------------------
1787 Bug-to-bug compatibility with some broken printers. On retransmit, try to send
1788 larger packets to work around bugs in certain TCP stacks. Can be turned off by
1789 setting it to zero.
1791 tcp_keepalive_probes
1792 --------------------
1794 Number of keep alive probes TCP sends out, until it decides that the
1795 connection is broken.
1797 tcp_keepalive_time
1798 ------------------
1800 How often TCP sends out keep alive messages, when keep alive is enabled. The
1801 default is 2 hours.
1803 tcp_syn_retries
1804 ---------------
1806 Number of times initial SYNs for a TCP connection attempt will be
1807 retransmitted. Should not be higher than 255. This is only the timeout for
1808 outgoing connections, for incoming connections the number of retransmits is
1809 defined by tcp_retries1.
1811 tcp_sack
1812 --------
1814 Enable select acknowledgments after RFC2018.
1816 tcp_timestamps
1817 --------------
1819 Enable timestamps as defined in RFC1323.
1821 tcp_stdurg
1822 ----------
1824 Enable the strict RFC793 interpretation of the TCP urgent pointer field. The
1825 default is to use the BSD compatible interpretation of the urgent pointer
1826 pointing to the first byte after the urgent data. The RFC793 interpretation is
1827 to have it point to the last byte of urgent data. Enabling this option may
1828 lead to interoperability problems. Disabled by default.
1830 tcp_syncookies
1831 --------------
1833 Only valid when the kernel was compiled with CONFIG_SYNCOOKIES. Send out
1834 syncookies when the syn backlog queue of a socket overflows. This is to ward
1835 off the common 'syn flood attack'. Disabled by default.
1837 Note that the concept of a socket backlog is abandoned. This means the peer
1838 may not receive reliable error messages from an over loaded server with
1839 syncookies enabled.
1841 tcp_window_scaling
1842 ------------------
1844 Enable window scaling as defined in RFC1323.
1846 tcp_fin_timeout
1847 ---------------
1849 The length of time in seconds it takes to receive a final FIN before the
1850 socket is always closed. This is strictly a violation of the TCP
1851 specification, but required to prevent denial-of-service attacks.
1853 tcp_max_ka_probes
1854 -----------------
1856 Indicates how many keep alive probes are sent per slow timer run. Should not
1857 be set too high to prevent bursts.
1859 tcp_max_syn_backlog
1860 -------------------
1862 Length of the per socket backlog queue. Since Linux 2.2 the backlog specified
1863 in listen(2) only specifies the length of the backlog queue of already
1864 established sockets. When more connection requests arrive Linux starts to drop
1865 packets. When syncookies are enabled the packets are still answered and the
1866 maximum queue is effectively ignored.
1868 tcp_retries1
1869 ------------
1871 Defines how often an answer to a TCP connection request is retransmitted
1872 before giving up.
1874 tcp_retries2
1875 ------------
1877 Defines how often a TCP packet is retransmitted before giving up.
1879 Interface specific settings
1880 ---------------------------
1882 In the directory /proc/sys/net/ipv4/conf you'll find one subdirectory for each
1883 interface the system knows about and one directory calls all. Changes in the
1884 all subdirectory affect all interfaces, whereas changes in the other
1885 subdirectories affect only one interface. All directories have the same
1886 entries:
1888 accept_redirects
1889 ----------------
1891 This switch decides if the kernel accepts ICMP redirect messages or not. The
1892 default is 'yes' if the kernel is configured for a regular host and 'no' for a
1893 router configuration.
1895 accept_source_route
1896 -------------------
1898 Should source routed packages be accepted or declined. The default is
1899 dependent on the kernel configuration. It's 'yes' for routers and 'no' for
1900 hosts.
1902 bootp_relay
1903 ~~~~~~~~~~~
1905 Accept packets with source address 0.b.c.d with destinations not to this host
1906 as local ones. It is supposed that a BOOTP relay daemon will catch and forward
1907 such packets.
1909 The default is 0, since this feature is not implemented yet (kernel version
1910 2.2.12).
1912 forwarding
1913 ----------
1915 Enable or disable IP forwarding on this interface.
1917 log_martians
1918 ------------
1920 Log packets with source addresses with no known route to kernel log.
1922 mc_forwarding
1923 -------------
1925 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE and a
1926 multicast routing daemon is required.
1928 proxy_arp
1929 ---------
1931 Does (1) or does not (0) perform proxy ARP.
1933 rp_filter
1934 ---------
1936 Integer value determines if a source validation should be made. 1 means yes, 0
1937 means no. Disabled by default, but local/broadcast address spoofing is always
1938 on.
1940 If you set this to 1 on a router that is the only connection for a network to
1941 the net, it will prevent spoofing attacks against your internal networks
1942 (external addresses can still be spoofed), without the need for additional
1943 firewall rules.
1945 secure_redirects
1946 ----------------
1948 Accept ICMP redirect messages only for gateways, listed in default gateway
1949 list. Enabled by default.
1951 shared_media
1952 ------------
1954 If it is not set the kernel does not assume that different subnets on this
1955 device can communicate directly. Default setting is 'yes'.
1957 send_redirects
1958 --------------
1960 Determines whether to send ICMP redirects to other hosts.
1962 Routing settings
1963 ----------------
1965 The directory /proc/sys/net/ipv4/route contains several file to control
1966 routing issues.
1968 error_burst and error_cost
1969 --------------------------
1971 These parameters are used to limit how many ICMP destination unreachable to
1972 send from the host in question. ICMP destination unreachable messages are
1973 sent when we cannot reach the next hop while trying to transmit a packet.
1974 It will also print some error messages to kernel logs if someone is ignoring
1975 our ICMP redirects. The higher the error_cost factor is, the fewer
1976 destination unreachable and error messages will be let through. Error_burst
1977 controls when destination unreachable messages and error messages will be
1978 dropped. The default settings limit warning messages to five every second.
1980 flush
1981 -----
1983 Writing to this file results in a flush of the routing cache.
1985 gc_elasticity, gc_interval, gc_min_interval_ms, gc_timeout, gc_thresh
1986 ---------------------------------------------------------------------
1988 Values to control the frequency and behavior of the garbage collection
1989 algorithm for the routing cache. gc_min_interval is deprecated and replaced
1990 by gc_min_interval_ms.
1993 max_size
1994 --------
1996 Maximum size of the routing cache. Old entries will be purged once the cache
1997 reached has this size.
1999 redirect_load, redirect_number
2000 ------------------------------
2002 Factors which determine if more ICPM redirects should be sent to a specific
2003 host. No redirects will be sent once the load limit or the maximum number of
2004 redirects has been reached.
2006 redirect_silence
2007 ----------------
2009 Timeout for redirects. After this period redirects will be sent again, even if
2010 this has been stopped, because the load or number limit has been reached.
2012 Network Neighbor handling
2013 -------------------------
2015 Settings about how to handle connections with direct neighbors (nodes attached
2016 to the same link) can be found in the directory /proc/sys/net/ipv4/neigh.
2018 As we saw it in the conf directory, there is a default subdirectory which
2019 holds the default values, and one directory for each interface. The contents
2020 of the directories are identical, with the single exception that the default
2021 settings contain additional options to set garbage collection parameters.
2023 In the interface directories you'll find the following entries:
2025 base_reachable_time, base_reachable_time_ms
2026 -------------------------------------------
2028 A base value used for computing the random reachable time value as specified
2029 in RFC2461.
2031 Expression of base_reachable_time, which is deprecated, is in seconds.
2032 Expression of base_reachable_time_ms is in milliseconds.
2034 retrans_time, retrans_time_ms
2035 -----------------------------
2037 The time between retransmitted Neighbor Solicitation messages.
2038 Used for address resolution and to determine if a neighbor is
2039 unreachable.
2041 Expression of retrans_time, which is deprecated, is in 1/100 seconds (for
2042 IPv4) or in jiffies (for IPv6).
2043 Expression of retrans_time_ms is in milliseconds.
2045 unres_qlen
2046 ----------
2048 Maximum queue length for a pending arp request - the number of packets which
2049 are accepted from other layers while the ARP address is still resolved.
2051 anycast_delay
2052 -------------
2054 Maximum for random delay of answers to neighbor solicitation messages in
2055 jiffies (1/100 sec). Not yet implemented (Linux does not have anycast support
2056 yet).
2058 ucast_solicit
2059 -------------
2061 Maximum number of retries for unicast solicitation.
2063 mcast_solicit
2064 -------------
2066 Maximum number of retries for multicast solicitation.
2068 delay_first_probe_time
2069 ----------------------
2071 Delay for the first time probe if the neighbor is reachable. (see
2072 gc_stale_time)
2074 locktime
2075 --------
2077 An ARP/neighbor entry is only replaced with a new one if the old is at least
2078 locktime old. This prevents ARP cache thrashing.
2080 proxy_delay
2081 -----------
2083 Maximum time (real time is random [0..proxytime]) before answering to an ARP
2084 request for which we have an proxy ARP entry. In some cases, this is used to
2085 prevent network flooding.
2087 proxy_qlen
2088 ----------
2090 Maximum queue length of the delayed proxy arp timer. (see proxy_delay).
2092 app_solicit
2093 ----------
2095 Determines the number of requests to send to the user level ARP daemon. Use 0
2096 to turn off.
2098 gc_stale_time
2099 -------------
2101 Determines how often to check for stale ARP entries. After an ARP entry is
2102 stale it will be resolved again (which is useful when an IP address migrates
2103 to another machine). When ucast_solicit is greater than 0 it first tries to
2104 send an ARP packet directly to the known host When that fails and
2105 mcast_solicit is greater than 0, an ARP request is broadcasted.
2107 2.9 Appletalk
2108 -------------
2110 The /proc/sys/net/appletalk directory holds the Appletalk configuration data
2111 when Appletalk is loaded. The configurable parameters are:
2113 aarp-expiry-time
2114 ----------------
2116 The amount of time we keep an ARP entry before expiring it. Used to age out
2117 old hosts.
2119 aarp-resolve-time
2120 -----------------
2122 The amount of time we will spend trying to resolve an Appletalk address.
2124 aarp-retransmit-limit
2125 ---------------------
2127 The number of times we will retransmit a query before giving up.
2129 aarp-tick-time
2130 --------------
2132 Controls the rate at which expires are checked.
2134 The directory /proc/net/appletalk holds the list of active Appletalk sockets
2135 on a machine.
2137 The fields indicate the DDP type, the local address (in network:node format)
2138 the remote address, the size of the transmit pending queue, the size of the
2139 received queue (bytes waiting for applications to read) the state and the uid
2140 owning the socket.
2142 /proc/net/atalk_iface lists all the interfaces configured for appletalk.It
2143 shows the name of the interface, its Appletalk address, the network range on
2144 that address (or network number for phase 1 networks), and the status of the
2145 interface.
2147 /proc/net/atalk_route lists each known network route. It lists the target
2148 (network) that the route leads to, the router (may be directly connected), the
2149 route flags, and the device the route is using.
2151 2.10 IPX
2152 --------
2154 The IPX protocol has no tunable values in proc/sys/net.
2156 The IPX protocol does, however, provide proc/net/ipx. This lists each IPX
2157 socket giving the local and remote addresses in Novell format (that is
2158 network:node:port). In accordance with the strange Novell tradition,
2159 everything but the port is in hex. Not_Connected is displayed for sockets that
2160 are not tied to a specific remote address. The Tx and Rx queue sizes indicate
2161 the number of bytes pending for transmission and reception. The state
2162 indicates the state the socket is in and the uid is the owning uid of the
2163 socket.
2165 The /proc/net/ipx_interface file lists all IPX interfaces. For each interface
2166 it gives the network number, the node number, and indicates if the network is
2167 the primary network. It also indicates which device it is bound to (or
2168 Internal for internal networks) and the Frame Type if appropriate. Linux
2169 supports 802.3, 802.2, 802.2 SNAP and DIX (Blue Book) ethernet framing for
2170 IPX.
2172 The /proc/net/ipx_route table holds a list of IPX routes. For each route it
2173 gives the destination network, the router node (or Directly) and the network
2174 address of the router (or Connected) for internal networks.
2176 2.11 /proc/sys/fs/mqueue - POSIX message queues filesystem
2177 ----------------------------------------------------------
2179 The "mqueue" filesystem provides the necessary kernel features to enable the
2180 creation of a user space library that implements the POSIX message queues
2181 API (as noted by the MSG tag in the POSIX 1003.1-2001 version of the System
2182 Interfaces specification.)
2184 The "mqueue" filesystem contains values for determining/setting the amount of
2185 resources used by the file system.
2187 /proc/sys/fs/mqueue/queues_max is a read/write file for setting/getting the
2188 maximum number of message queues allowed on the system.
2190 /proc/sys/fs/mqueue/msg_max is a read/write file for setting/getting the
2191 maximum number of messages in a queue value. In fact it is the limiting value
2192 for another (user) limit which is set in mq_open invocation. This attribute of
2193 a queue must be less or equal then msg_max.
2195 /proc/sys/fs/mqueue/msgsize_max is a read/write file for setting/getting the
2196 maximum message size value (it is every message queue's attribute set during
2197 its creation).
2199 2.12 /proc/<pid>/oom_adj - Adjust the oom-killer score
2200 ------------------------------------------------------
2202 This file can be used to adjust the score used to select which processes
2203 should be killed in an out-of-memory situation. Giving it a high score will
2204 increase the likelihood of this process being killed by the oom-killer. Valid
2205 values are in the range -16 to +15, plus the special value -17, which disables
2206 oom-killing altogether for this process.
2208 2.13 /proc/<pid>/oom_score - Display current oom-killer score
2209 -------------------------------------------------------------
2211 ------------------------------------------------------------------------------
2212 This file can be used to check the current score used by the oom-killer is for
2213 any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
2214 process should be killed in an out-of-memory situation.
2216 ------------------------------------------------------------------------------
2217 Summary
2218 ------------------------------------------------------------------------------
2219 Certain aspects of kernel behavior can be modified at runtime, without the
2220 need to recompile the kernel, or even to reboot the system. The files in the
2221 /proc/sys tree can not only be read, but also modified. You can use the echo
2222 command to write value into these files, thereby changing the default settings
2223 of the kernel.
2224 ------------------------------------------------------------------------------
2226 2.14 /proc/<pid>/io - Display the IO accounting fields
2227 -------------------------------------------------------
2229 This file contains IO statistics for each running process
2231 Example
2232 -------
2234 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
2235 [1] 3828
2237 test:/tmp # cat /proc/3828/io
2238 rchar: 323934931
2239 wchar: 323929600
2240 syscr: 632687
2241 syscw: 632675
2242 read_bytes: 0
2243 write_bytes: 323932160
2244 cancelled_write_bytes: 0
2247 Description
2248 -----------
2250 rchar
2251 -----
2253 I/O counter: chars read
2254 The number of bytes which this task has caused to be read from storage. This
2255 is simply the sum of bytes which this process passed to read() and pread().
2256 It includes things like tty IO and it is unaffected by whether or not actual
2257 physical disk IO was required (the read might have been satisfied from
2258 pagecache)
2261 wchar
2262 -----
2264 I/O counter: chars written
2265 The number of bytes which this task has caused, or shall cause to be written
2266 to disk. Similar caveats apply here as with rchar.
2269 syscr
2270 -----
2272 I/O counter: read syscalls
2273 Attempt to count the number of read I/O operations, i.e. syscalls like read()
2274 and pread().
2277 syscw
2278 -----
2280 I/O counter: write syscalls
2281 Attempt to count the number of write I/O operations, i.e. syscalls like
2282 write() and pwrite().
2285 read_bytes
2286 ----------
2288 I/O counter: bytes read
2289 Attempt to count the number of bytes which this process really did cause to
2290 be fetched from the storage layer. Done at the submit_bio() level, so it is
2291 accurate for block-backed filesystems. <please add status regarding NFS and
2292 CIFS at a later time>
2295 write_bytes
2296 -----------
2298 I/O counter: bytes written
2299 Attempt to count the number of bytes which this process caused to be sent to
2300 the storage layer. This is done at page-dirtying time.
2303 cancelled_write_bytes
2304 ---------------------
2306 The big inaccuracy here is truncate. If a process writes 1MB to a file and
2307 then deletes the file, it will in fact perform no writeout. But it will have
2308 been accounted as having caused 1MB of write.
2309 In other words: The number of bytes which this process caused to not happen,
2310 by truncating pagecache. A task can cause "negative" IO too. If this task
2311 truncates some dirty pagecache, some IO which another task has been accounted
2312 for (in it's write_bytes) will not be happening. We _could_ just subtract that
2313 from the truncating task's write_bytes, but there is information loss in doing
2314 that.
2317 Note
2318 ----
2320 At its current implementation state, this is a bit racy on 32-bit machines: if
2321 process A reads process B's /proc/pid/io while process B is updating one of
2322 those 64-bit counters, process A could see an intermediate result.
2325 More information about this can be found within the taskstats documentation in
2326 Documentation/accounting.
2328 2.15 /proc/<pid>/coredump_filter - Core dump filtering settings
2329 ---------------------------------------------------------------
2330 When a process is dumped, all anonymous memory is written to a core file as
2331 long as the size of the core file isn't limited. But sometimes we don't want
2332 to dump some memory segments, for example, huge shared memory. Conversely,
2333 sometimes we want to save file-backed memory segments into a core file, not
2334 only the individual files.
2336 /proc/<pid>/coredump_filter allows you to customize which memory segments
2337 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
2338 of memory types. If a bit of the bitmask is set, memory segments of the
2339 corresponding memory type are dumped, otherwise they are not dumped.
2341 The following 4 memory types are supported:
2342 - (bit 0) anonymous private memory
2343 - (bit 1) anonymous shared memory
2344 - (bit 2) file-backed private memory
2345 - (bit 3) file-backed shared memory
2347 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
2348 are always dumped regardless of the bitmask status.
2350 Default value of coredump_filter is 0x3; this means all anonymous memory
2351 segments are dumped.
2353 If you don't want to dump all shared memory segments attached to pid 1234,
2354 write 1 to the process's proc file.
2356 $ echo 0x1 > /proc/1234/coredump_filter
2358 When a new process is created, the process inherits the bitmask status from its
2359 parent. It is useful to set up coredump_filter before the program runs.
2360 For example:
2362 $ echo 0x7 > /proc/self/coredump_filter
2363 $ ./some_program
2365 2.16 /proc/<pid>/mountinfo - Information about mounts
2366 --------------------------------------------------------
2368 This file contains lines of the form:
2370 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
2371 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
2373 (1) mount ID: unique identifier of the mount (may be reused after umount)
2374 (2) parent ID: ID of parent (or of self for the top of the mount tree)
2375 (3) major:minor: value of st_dev for files on filesystem
2376 (4) root: root of the mount within the filesystem
2377 (5) mount point: mount point relative to the process's root
2378 (6) mount options: per mount options
2379 (7) optional fields: zero or more fields of the form "tag[:value]"
2380 (8) separator: marks the end of the optional fields
2381 (9) filesystem type: name of filesystem of the form "type[.subtype]"
2382 (10) mount source: filesystem specific information or "none"
2383 (11) super options: per super block options
2385 Parsers should ignore all unrecognised optional fields. Currently the
2386 possible optional fields are:
2388 shared:X mount is shared in peer group X
2389 master:X mount is slave to peer group X
2390 propagate_from:X mount is slave and receives propagation from peer group X (*)
2391 unbindable mount is unbindable
2393 (*) X is the closest dominant peer group under the process's root. If
2394 X is the immediate master of the mount, or if there's no dominant peer
2395 group under the same root, then only the "master:X" field is present
2396 and not the "propagate_from:X" field.
2398 For more information on mount propagation see:
2400 Documentation/filesystems/sharedsubtree.txt
2402 ------------------------------------------------------------------------------