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 ------------------------------------------------------------------------------
17 0.1 Introduction/Credits
20 1 Collecting System Information
21 1.1 Process-Specific Subdirectories
23 1.3 IDE devices in /proc/ide
24 1.4 Networking info in /proc/net
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
41 2.11 /proc/sys/fs/mqueue - POSIX message queues filesystem
43 ------------------------------------------------------------------------------
45 ------------------------------------------------------------------------------
47 0.1 Introduction/Credits
48 ------------------------
50 This documentation is part of a soon (or so we hope) to be released book on
51 the SuSE Linux distribution. As there is no complete documentation for the
52 /proc file system and we've used many freely available sources to write these
53 chapters, it seems only fair to give the work back to the Linux community.
54 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
55 afraid it's still far from complete, but we hope it will be useful. As far as
56 we know, it is the first 'all-in-one' document about the /proc file system. It
57 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
58 SPARC, APX, etc., features, you probably won't find what you are looking for.
59 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
60 additions and patches are welcome and will be added to this document if you
63 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
64 other people for help compiling this documentation. We'd also like to extend a
65 special thank you to Andi Kleen for documentation, which we relied on heavily
66 to create this document, as well as the additional information he provided.
67 Thanks to everybody else who contributed source or docs to the Linux kernel
68 and helped create a great piece of software... :)
70 If you have any comments, corrections or additions, please don't hesitate to
71 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
74 The latest version of this document is available online at
75 http://skaro.nightcrawler.com/~bb/Docs/Proc as HTML version.
77 If the above direction does not works for you, ypu could try the kernel
78 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
79 comandante@zaralinux.com.
84 We don't guarantee the correctness of this document, and if you come to us
85 complaining about how you screwed up your system because of incorrect
86 documentation, we won't feel responsible...
88 ------------------------------------------------------------------------------
89 CHAPTER 1: COLLECTING SYSTEM INFORMATION
90 ------------------------------------------------------------------------------
92 ------------------------------------------------------------------------------
94 ------------------------------------------------------------------------------
95 * Investigating the properties of the pseudo file system /proc and its
96 ability to provide information on the running Linux system
97 * Examining /proc's structure
98 * Uncovering various information about the kernel and the processes running
100 ------------------------------------------------------------------------------
103 The proc file system acts as an interface to internal data structures in the
104 kernel. It can be used to obtain information about the system and to change
105 certain kernel parameters at runtime (sysctl).
107 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
108 show you how you can use /proc/sys to change settings.
110 1.1 Process-Specific Subdirectories
111 -----------------------------------
113 The directory /proc contains (among other things) one subdirectory for each
114 process running on the system, which is named after the process ID (PID).
116 The link self points to the process reading the file system. Each process
117 subdirectory has the entries listed in Table 1-1.
120 Table 1-1: Process specific entries in /proc
121 ..............................................................................
123 cmdline Command line arguments
124 cpu Current and last cpu in wich it was executed (2.4)(smp)
125 cwd Link to the current working directory
126 environ Values of environment variables
127 exe Link to the executable of this process
128 fd Directory, which contains all file descriptors
129 maps Memory maps to executables and library files (2.4)
130 mem Memory held by this process
131 root Link to the root directory of this process
133 statm Process memory status information
134 status Process status in human readable form
135 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
136 ..............................................................................
138 For example, to get the status information of a process, all you have to do is
139 read the file /proc/PID/status:
141 >cat /proc/self/status
157 SigPnd: 0000000000000000
158 SigBlk: 0000000000000000
159 SigIgn: 0000000000000000
160 SigCgt: 0000000000000000
161 CapInh: 00000000fffffeff
162 CapPrm: 0000000000000000
163 CapEff: 0000000000000000
166 This shows you nearly the same information you would get if you viewed it with
167 the ps command. In fact, ps uses the proc file system to obtain its
168 information. The statm file contains more detailed information about the
169 process memory usage. Its seven fields are explained in Table 1-2.
172 Table 1-2: Contents of the statm files
173 ..............................................................................
175 size total program size
176 resident size of memory portions
177 shared number of pages that are shared
178 trs number of pages that are 'code'
179 drs number of pages of data/stack
180 lrs number of pages of library
181 dt number of dirty pages
182 ..............................................................................
187 Similar to the process entries, the kernel data files give information about
188 the running kernel. The files used to obtain this information are contained in
189 /proc and are listed in Table 1-3. Not all of these will be present in your
190 system. It depends on the kernel configuration and the loaded modules, which
191 files are there, and which are missing.
193 Table 1-3: Kernel info in /proc
194 ..............................................................................
196 apm Advanced power management info
197 buddyinfo Kernel memory allocator information (see text) (2.5)
198 bus Directory containing bus specific information
199 cmdline Kernel command line
200 cpuinfo Info about the CPU
201 devices Available devices (block and character)
202 dma Used DMS channels
203 filesystems Supported filesystems
204 driver Various drivers grouped here, currently rtc (2.4)
205 execdomains Execdomains, related to security (2.4)
206 fb Frame Buffer devices (2.4)
207 fs File system parameters, currently nfs/exports (2.4)
208 ide Directory containing info about the IDE subsystem
209 interrupts Interrupt usage
210 iomem Memory map (2.4)
211 ioports I/O port usage
212 irq Masks for irq to cpu affinity (2.4)(smp?)
213 isapnp ISA PnP (Plug&Play) Info (2.4)
214 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
216 ksyms Kernel symbol table
217 loadavg Load average of last 1, 5 & 15 minutes
221 modules List of loaded modules
222 mounts Mounted filesystems
223 net Networking info (see text)
224 partitions Table of partitions known to the system
225 pci Depreciated info of PCI bus (new way -> /proc/bus/pci/,
226 decoupled by lspci (2.4)
228 scsi SCSI info (see text)
229 slabinfo Slab pool info
230 stat Overall statistics
231 swaps Swap space utilization
233 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
234 tty Info of tty drivers
236 version Kernel version
237 video bttv info of video resources (2.4)
238 ..............................................................................
240 You can, for example, check which interrupts are currently in use and what
241 they are used for by looking in the file /proc/interrupts:
243 > cat /proc/interrupts
245 0: 8728810 XT-PIC timer
246 1: 895 XT-PIC keyboard
248 3: 531695 XT-PIC aha152x
249 4: 2014133 XT-PIC serial
250 5: 44401 XT-PIC pcnet_cs
253 12: 182918 XT-PIC PS/2 Mouse
255 14: 1232265 XT-PIC ide0
259 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
260 output of a SMP machine):
262 > cat /proc/interrupts
265 0: 1243498 1214548 IO-APIC-edge timer
266 1: 8949 8958 IO-APIC-edge keyboard
267 2: 0 0 XT-PIC cascade
268 5: 11286 10161 IO-APIC-edge soundblaster
269 8: 1 0 IO-APIC-edge rtc
270 9: 27422 27407 IO-APIC-edge 3c503
271 12: 113645 113873 IO-APIC-edge PS/2 Mouse
273 14: 22491 24012 IO-APIC-edge ide0
274 15: 2183 2415 IO-APIC-edge ide1
275 17: 30564 30414 IO-APIC-level eth0
276 18: 177 164 IO-APIC-level bttv
281 NMI is incremented in this case because every timer interrupt generates a NMI
282 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
284 LOC is the local interrupt counter of the internal APIC of every CPU.
286 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
287 connects the CPUs in a SMP system. This means that an error has been detected,
288 the IO-APIC automatically retry the transmission, so it should not be a big
289 problem, but you should read the SMP-FAQ.
291 In this context it could be interesting to note the new irq directory in 2.4.
292 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
293 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
294 irq subdir is one subdir for each IRQ, and one file; prof_cpu_mask
298 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
299 1 11 13 15 17 19 3 5 7 9
303 The contents of the prof_cpu_mask file and each smp_affinity file for each IRQ
304 is the same by default:
306 > cat /proc/irq/0/smp_affinity
309 It's a bitmask, in wich you can specify wich CPUs can handle the IRQ, you can
312 > echo 1 > /proc/irq/prof_cpu_mask
314 This means that only the first CPU will handle the IRQ, but you can also echo 5
315 wich means that only the first and fourth CPU can handle the IRQ.
317 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
318 between all the CPUs which are allowed to handle it. As usual the kernel has
319 more info than you and does a better job than you, so the defaults are the
320 best choice for almost everyone.
322 There are three more important subdirectories in /proc: net, scsi, and sys.
323 The general rule is that the contents, or even the existence of these
324 directories, depend on your kernel configuration. If SCSI is not enabled, the
325 directory scsi may not exist. The same is true with the net, which is there
326 only when networking support is present in the running kernel.
328 The slabinfo file gives information about memory usage at the slab level.
329 Linux uses slab pools for memory management above page level in version 2.2.
330 Commonly used objects have their own slab pool (such as network buffers,
331 directory cache, and so on).
333 ..............................................................................
335 > cat /proc/buddyinfo
337 Node 0, zone DMA 0 4 5 4 4 3 ...
338 Node 0, zone Normal 1 0 0 1 101 8 ...
339 Node 0, zone HighMem 2 0 0 1 1 0 ...
341 Memory fragmentation is a problem under some workloads, and buddyinfo is a
342 useful tool for helping diagnose these problems. Buddyinfo will give you a
343 clue as to how big an area you can safely allocate, or why a previous
346 Each column represents the number of pages of a certain order which are
347 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
348 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
349 available in ZONE_NORMAL, etc...
352 1.3 IDE devices in /proc/ide
353 ----------------------------
355 The subdirectory /proc/ide contains information about all IDE devices of which
356 the kernel is aware. There is one subdirectory for each IDE controller, the
357 file drivers and a link for each IDE device, pointing to the device directory
358 in the controller specific subtree.
360 The file drivers contains general information about the drivers used for the
363 > cat /proc/ide/drivers
364 ide-cdrom version 4.53
365 ide-disk version 1.08
367 ..............................................................................
371 Provides information about distribution and utilization of memory. This
372 varies by architecture and compile options. The following is from a
373 16GB PIII, which has highmem enabled. You may not have all of these fields.
378 MemTotal: 16344972 kB
385 HighTotal: 15597528 kB
386 HighFree: 13629632 kB
395 Committed_AS: 1576424 kB
398 VmallocTotal: 112216 kB
400 VmallocChunk: 111088 kB
402 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
403 bits and the kernel binary code)
404 MemFree: The sum of LowFree+HighFree
405 Buffers: Relatively temporary storage for raw disk blocks
406 shouldn't get tremendously large (20MB or so)
407 Cached: in-memory cache for files read from the disk (the
408 pagecache). Doesn't include SwapCached
409 SwapCached: Memory that once was swapped out, is swapped back in but
410 still also is in the swapfile (if memory is needed it
411 doesn't need to be swapped out AGAIN because it is already
412 in the swapfile. This saves I/O)
413 Active: Memory that has been used more recently and usually not
414 reclaimed unless absolutely necessary.
415 Inactive: Memory which has been less recently used. It is more
416 eligible to be reclaimed for other purposes
418 HighFree: Highmem is all memory above ~860MB of physical memory
419 Highmem areas are for use by userspace programs, or
420 for the pagecache. The kernel must use tricks to access
421 this memory, making it slower to access than lowmem.
423 LowFree: Lowmem is memory which can be used for everything that
424 highmem can be used for, but it is also availble for the
425 kernel's use for its own data structures. Among many
426 other things, it is where everything from the Slab is
427 allocated. Bad things happen when you're out of lowmem.
428 SwapTotal: total amount of swap space available
429 SwapFree: Memory which has been evicted from RAM, and is temporarily
431 Dirty: Memory which is waiting to get written back to the disk
432 Writeback: Memory which is actively being written back to the disk
433 Mapped: files which have been mmaped, such as libraries
434 Slab: in-kernel data structures cache
435 Committed_AS: An estimate of how much RAM you would need to make a
436 99.99% guarantee that there never is OOM (out of memory)
437 for this workload. Normally the kernel will overcommit
438 memory. That means, say you do a 1GB malloc, nothing
439 happens, really. Only when you start USING that malloc
440 memory you will get real memory on demand, and just as
441 much as you use. So you sort of take a mortgage and hope
442 the bank doesn't go bust. Other cases might include when
443 you mmap a file that's shared only when you write to it
444 and you get a private copy of that data. While it normally
445 is shared between processes. The Committed_AS is a
446 guesstimate of how much RAM/swap you would need
448 PageTables: amount of memory dedicated to the lowest level of page
450 ReverseMaps: number of reverse mappings performed
451 VmallocTotal: total size of vmalloc memory area
452 VmallocUsed: amount of vmalloc area which is used
453 VmallocChunk: largest contigious block of vmalloc area which is free
455 More detailed information can be found in the controller specific
456 subdirectories. These are named ide0, ide1 and so on. Each of these
457 directories contains the files shown in table 1-4.
460 Table 1-4: IDE controller info in /proc/ide/ide?
461 ..............................................................................
463 channel IDE channel (0 or 1)
464 config Configuration (only for PCI/IDE bridge)
466 model Type/Chipset of IDE controller
467 ..............................................................................
469 Each device connected to a controller has a separate subdirectory in the
470 controllers directory. The files listed in table 1-5 are contained in these
474 Table 1-5: IDE device information
475 ..............................................................................
478 capacity Capacity of the medium (in 512Byte blocks)
479 driver driver and version
480 geometry physical and logical geometry
481 identify device identify block
483 model device identifier
484 settings device setup
485 smart_thresholds IDE disk management thresholds
486 smart_values IDE disk management values
487 ..............................................................................
489 The most interesting file is settings. This file contains a nice overview of
490 the drive parameters:
492 # cat /proc/ide/ide0/hda/settings
493 name value min max mode
494 ---- ----- --- --- ----
495 bios_cyl 526 0 65535 rw
496 bios_head 255 0 255 rw
498 breada_readahead 4 0 127 rw
500 file_readahead 72 0 2097151 rw
502 keepsettings 0 0 1 rw
503 max_kb_per_request 122 1 127 rw
507 pio_mode write-only 0 255 w
513 1.4 Networking info in /proc/net
514 --------------------------------
516 The subdirectory /proc/net follows the usual pattern. Table 1-6 shows the
517 additional values you get for IP version 6 if you configure the kernel to
518 support this. Table 1-7 lists the files and their meaning.
521 Table 1-6: IPv6 info in /proc/net
522 ..............................................................................
524 udp6 UDP sockets (IPv6)
525 tcp6 TCP sockets (IPv6)
526 raw6 Raw device statistics (IPv6)
527 igmp6 IP multicast addresses, which this host joined (IPv6)
528 if_inet6 List of IPv6 interface addresses
529 ipv6_route Kernel routing table for IPv6
530 rt6_stats Global IPv6 routing tables statistics
531 sockstat6 Socket statistics (IPv6)
532 snmp6 Snmp data (IPv6)
533 ..............................................................................
536 Table 1-7: Network info in /proc/net
537 ..............................................................................
540 dev network devices with statistics
541 dev_mcast the Layer2 multicast groups a device is listening too
542 (interface index, label, number of references, number of bound
544 dev_stat network device status
545 ip_fwchains Firewall chain linkage
546 ip_fwnames Firewall chain names
547 ip_masq Directory containing the masquerading tables
548 ip_masquerade Major masquerading table
549 netstat Network statistics
550 raw raw device statistics
551 route Kernel routing table
552 rpc Directory containing rpc info
553 rt_cache Routing cache
555 sockstat Socket statistics
557 tr_rif Token ring RIF routing table
559 unix UNIX domain sockets
560 wireless Wireless interface data (Wavelan etc)
561 igmp IP multicast addresses, which this host joined
562 psched Global packet scheduler parameters.
563 netlink List of PF_NETLINK sockets
564 ip_mr_vifs List of multicast virtual interfaces
565 ip_mr_cache List of multicast routing cache
566 ..............................................................................
568 You can use this information to see which network devices are available in
569 your system and how much traffic was routed over those devices:
573 face |bytes packets errs drop fifo frame compressed multicast|[...
574 lo: 908188 5596 0 0 0 0 0 0 [...
575 ppp0:15475140 20721 410 0 0 410 0 0 [...
576 eth0: 614530 7085 0 0 0 0 0 1 [...
579 ...] bytes packets errs drop fifo colls carrier compressed
580 ...] 908188 5596 0 0 0 0 0 0
581 ...] 1375103 17405 0 0 0 0 0 0
582 ...] 1703981 5535 0 0 0 3 0 0
584 In addition, each Channel Bond interface has it's own directory. For
585 example, the bond0 device will have a directory called /proc/net/bond0/.
586 It will contain information that is specific to that bond, such as the
587 current slaves of the bond, the link status of the slaves, and how
588 many times the slaves link has failed.
593 If you have a SCSI host adapter in your system, you'll find a subdirectory
594 named after the driver for this adapter in /proc/scsi. You'll also see a list
595 of all recognized SCSI devices in /proc/scsi:
599 Host: scsi0 Channel: 00 Id: 00 Lun: 00
600 Vendor: IBM Model: DGHS09U Rev: 03E0
601 Type: Direct-Access ANSI SCSI revision: 03
602 Host: scsi0 Channel: 00 Id: 06 Lun: 00
603 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
604 Type: CD-ROM ANSI SCSI revision: 02
607 The directory named after the driver has one file for each adapter found in
608 the system. These files contain information about the controller, including
609 the used IRQ and the IO address range. The amount of information shown is
610 dependent on the adapter you use. The example shows the output for an Adaptec
611 AHA-2940 SCSI adapter:
613 > cat /proc/scsi/aic7xxx/0
615 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
617 TCQ Enabled By Default : Disabled
618 AIC7XXX_PROC_STATS : Disabled
619 AIC7XXX_RESET_DELAY : 5
620 Adapter Configuration:
621 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
622 Ultra Wide Controller
623 PCI MMAPed I/O Base: 0xeb001000
624 Adapter SEEPROM Config: SEEPROM found and used.
625 Adaptec SCSI BIOS: Enabled
627 SCBs: Active 0, Max Active 2,
628 Allocated 15, HW 16, Page 255
630 BIOS Control Word: 0x18b6
631 Adapter Control Word: 0x005b
632 Extended Translation: Enabled
633 Disconnect Enable Flags: 0xffff
634 Ultra Enable Flags: 0x0001
635 Tag Queue Enable Flags: 0x0000
636 Ordered Queue Tag Flags: 0x0000
637 Default Tag Queue Depth: 8
638 Tagged Queue By Device array for aic7xxx host instance 0:
639 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
640 Actual queue depth per device for aic7xxx host instance 0:
641 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
644 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
645 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
646 Total transfers 160151 (74577 reads and 85574 writes)
648 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
649 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
650 Total transfers 0 (0 reads and 0 writes)
653 1.6 Parallel port info in /proc/parport
654 ---------------------------------------
656 The directory /proc/parport contains information about the parallel ports of
657 your system. It has one subdirectory for each port, named after the port
660 These directories contain the four files shown in Table 1-8.
663 Table 1-8: Files in /proc/parport
664 ..............................................................................
666 autoprobe Any IEEE-1284 device ID information that has been acquired.
667 devices list of the device drivers using that port. A + will appear by the
668 name of the device currently using the port (it might not appear
670 hardware Parallel port's base address, IRQ line and DMA channel.
671 irq IRQ that parport is using for that port. This is in a separate
672 file to allow you to alter it by writing a new value in (IRQ
674 ..............................................................................
676 1.7 TTY info in /proc/tty
677 -------------------------
679 Information about the available and actually used tty's can be found in the
680 directory /proc/tty.You'll find entries for drivers and line disciplines in
681 this directory, as shown in Table 1-9.
684 Table 1-9: Files in /proc/tty
685 ..............................................................................
687 drivers list of drivers and their usage
688 ldiscs registered line disciplines
689 driver/serial usage statistic and status of single tty lines
690 ..............................................................................
692 To see which tty's are currently in use, you can simply look into the file
695 > cat /proc/tty/drivers
696 pty_slave /dev/pts 136 0-255 pty:slave
697 pty_master /dev/ptm 128 0-255 pty:master
698 pty_slave /dev/ttyp 3 0-255 pty:slave
699 pty_master /dev/pty 2 0-255 pty:master
700 serial /dev/cua 5 64-67 serial:callout
701 serial /dev/ttyS 4 64-67 serial
702 /dev/tty0 /dev/tty0 4 0 system:vtmaster
703 /dev/ptmx /dev/ptmx 5 2 system
704 /dev/console /dev/console 5 1 system:console
705 /dev/tty /dev/tty 5 0 system:/dev/tty
706 unknown /dev/tty 4 1-63 console
709 1.8 Miscellaneous kernel statistics in /proc/stat
710 -------------------------------------------------
712 Various pieces of information about kernel activity are available in the
713 /proc/stat file. All of the numbers reported in this file are aggregates
714 since the system first booted. For a quick look, simply cat the file:
717 cpu 2255 34 2290 22625563 6290 127 456
718 cpu0 1132 34 1441 11311718 3675 127 438
719 cpu1 1123 0 849 11313845 2614 0 18
720 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
727 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
728 lines. These numbers identify the amount of time the CPU has spent performing
729 different kinds of work. Time units are in USER_HZ (typically hundredths of a
730 second). The meanings of the columns are as follows, from left to right:
732 - user: normal processes executing in user mode
733 - nice: niced processes executing in user mode
734 - system: processes executing in kernel mode
735 - idle: twiddling thumbs
736 - iowait: waiting for I/O to complete
737 - irq: servicing interrupts
738 - softirq: servicing softirqs
740 The "intr" line gives counts of interrupts serviced since boot time, for each
741 of the possible system interrupts. The first column is the total of all
742 interrupts serviced; each subsequent column is the total for that particular
745 The "ctxt" line gives the total number of context switches across all CPUs.
747 The "btime" line gives the time at which the system booted, in seconds since
750 The "processes" line gives the number of processes and threads created, which
751 includes (but is not limited to) those created by calls to the fork() and
752 clone() system calls.
754 The "procs_running" line gives the number of processes currently running on
757 The "procs_blocked" line gives the number of processes currently blocked,
758 waiting for I/O to complete.
761 ------------------------------------------------------------------------------
763 ------------------------------------------------------------------------------
764 The /proc file system serves information about the running system. It not only
765 allows access to process data but also allows you to request the kernel status
766 by reading files in the hierarchy.
768 The directory structure of /proc reflects the types of information and makes
769 it easy, if not obvious, where to look for specific data.
770 ------------------------------------------------------------------------------
772 ------------------------------------------------------------------------------
773 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
774 ------------------------------------------------------------------------------
776 ------------------------------------------------------------------------------
778 ------------------------------------------------------------------------------
779 * Modifying kernel parameters by writing into files found in /proc/sys
780 * Exploring the files which modify certain parameters
781 * Review of the /proc/sys file tree
782 ------------------------------------------------------------------------------
785 A very interesting part of /proc is the directory /proc/sys. This is not only
786 a source of information, it also allows you to change parameters within the
787 kernel. Be very careful when attempting this. You can optimize your system,
788 but you can also cause it to crash. Never alter kernel parameters on a
789 production system. Set up a development machine and test to make sure that
790 everything works the way you want it to. You may have no alternative but to
791 reboot the machine once an error has been made.
793 To change a value, simply echo the new value into the file. An example is
794 given below in the section on the file system data. You need to be root to do
795 this. You can create your own boot script to perform this every time your
798 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
799 general things in the operation of the Linux kernel. Since some of the files
800 can inadvertently disrupt your system, it is advisable to read both
801 documentation and source before actually making adjustments. In any case, be
802 very careful when writing to any of these files. The entries in /proc may
803 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
804 review the kernel documentation in the directory /usr/src/linux/Documentation.
805 This chapter is heavily based on the documentation included in the pre 2.2
806 kernels, and became part of it in version 2.2.1 of the Linux kernel.
808 2.1 /proc/sys/fs - File system data
809 -----------------------------------
811 This subdirectory contains specific file system, file handle, inode, dentry
812 and quota information.
814 Currently, these files are in /proc/sys/fs:
819 Status of the directory cache. Since directory entries are dynamically
820 allocated and deallocated, this file indicates the current status. It holds
821 six values, in which the last two are not used and are always zero. The others
822 are listed in table 2-1.
825 Table 2-1: Status files of the directory cache
826 ..............................................................................
828 nr_dentry Almost always zero
829 nr_unused Number of unused cache entries
831 in seconds after the entry may be reclaimed, when memory is short
832 want_pages internally
833 ..............................................................................
835 dquot-nr and dquot-max
836 ----------------------
838 The file dquot-max shows the maximum number of cached disk quota entries.
840 The file dquot-nr shows the number of allocated disk quota entries and the
841 number of free disk quota entries.
843 If the number of available cached disk quotas is very low and you have a large
844 number of simultaneous system users, you might want to raise the limit.
849 The kernel allocates file handles dynamically, but doesn't free them again at
852 The value in file-max denotes the maximum number of file handles that the
853 Linux kernel will allocate. When you get a lot of error messages about running
854 out of file handles, you might want to raise this limit. The default value is
855 10% of RAM in kilobytes. To change it, just write the new number into the
858 # cat /proc/sys/fs/file-max
860 # echo 8192 > /proc/sys/fs/file-max
861 # cat /proc/sys/fs/file-max
865 This method of revision is useful for all customizable parameters of the
866 kernel - simply echo the new value to the corresponding file.
868 Historically, the three values in file-nr denoted the number of allocated file
869 handles, the number of allocated but unused file handles, and the maximum
870 number of file handles. Linux 2.6 always reports 0 as the number of free file
871 handles -- this is not an error, it just means that the number of allocated
872 file handles exactly matches the number of used file handles.
874 Attempts to allocate more file descriptors than file-max are reported with
875 printk, look for "VFS: file-max limit <number> reached".
877 inode-state and inode-nr
878 ------------------------
880 The file inode-nr contains the first two items from inode-state, so we'll skip
883 inode-state contains two actual numbers and five dummy values. The numbers
884 are nr_inodes and nr_free_inodes (in order of appearance).
889 Denotes the number of inodes the system has allocated. This number will
890 grow and shrink dynamically.
895 Represents the number of free inodes. Ie. The number of inuse inodes is
896 (nr_inodes - nr_free_inodes).
898 super-nr and super-max
899 ----------------------
901 Again, super block structures are allocated by the kernel, but not freed. The
902 file super-max contains the maximum number of super block handlers, where
903 super-nr shows the number of currently allocated ones.
905 Every mounted file system needs a super block, so if you plan to mount lots of
906 file systems, you may want to increase these numbers.
908 aio-nr and aio-max-nr
909 ---------------------
911 aio-nr is the running total of the number of events specified on the
912 io_setup system call for all currently active aio contexts. If aio-nr
913 reaches aio-max-nr then io_setup will fail with EAGAIN. Note that
914 raising aio-max-nr does not result in the pre-allocation or re-sizing
915 of any kernel data structures.
917 2.2 /proc/sys/fs/binfmt_misc - Miscellaneous binary formats
918 -----------------------------------------------------------
920 Besides these files, there is the subdirectory /proc/sys/fs/binfmt_misc. This
921 handles the kernel support for miscellaneous binary formats.
923 Binfmt_misc provides the ability to register additional binary formats to the
924 Kernel without compiling an additional module/kernel. Therefore, binfmt_misc
925 needs to know magic numbers at the beginning or the filename extension of the
928 It works by maintaining a linked list of structs that contain a description of
929 a binary format, including a magic with size (or the filename extension),
930 offset and mask, and the interpreter name. On request it invokes the given
931 interpreter with the original program as argument, as binfmt_java and
932 binfmt_em86 and binfmt_mz do. Since binfmt_misc does not define any default
933 binary-formats, you have to register an additional binary-format.
935 There are two general files in binfmt_misc and one file per registered format.
936 The two general files are register and status.
938 Registering a new binary format
939 -------------------------------
941 To register a new binary format you have to issue the command
943 echo :name:type:offset:magic:mask:interpreter: > /proc/sys/fs/binfmt_misc/register
947 with appropriate name (the name for the /proc-dir entry), offset (defaults to
948 0, if omitted), magic, mask (which can be omitted, defaults to all 0xff) and
949 last but not least, the interpreter that is to be invoked (for example and
950 testing /bin/echo). Type can be M for usual magic matching or E for filename
951 extension matching (give extension in place of magic).
953 Check or reset the status of the binary format handler
954 ------------------------------------------------------
956 If you do a cat on the file /proc/sys/fs/binfmt_misc/status, you will get the
957 current status (enabled/disabled) of binfmt_misc. Change the status by echoing
958 0 (disables) or 1 (enables) or -1 (caution: this clears all previously
959 registered binary formats) to status. For example echo 0 > status to disable
960 binfmt_misc (temporarily).
962 Status of a single handler
963 --------------------------
965 Each registered handler has an entry in /proc/sys/fs/binfmt_misc. These files
966 perform the same function as status, but their scope is limited to the actual
967 binary format. By cating this file, you also receive all related information
968 about the interpreter/magic of the binfmt.
970 Example usage of binfmt_misc (emulate binfmt_java)
971 --------------------------------------------------
973 cd /proc/sys/fs/binfmt_misc
974 echo ':Java:M::\xca\xfe\xba\xbe::/usr/local/java/bin/javawrapper:' > register
975 echo ':HTML:E::html::/usr/local/java/bin/appletviewer:' > register
976 echo ':Applet:M::<!--applet::/usr/local/java/bin/appletviewer:' > register
977 echo ':DEXE:M::\x0eDEX::/usr/bin/dosexec:' > register
980 These four lines add support for Java executables and Java applets (like
981 binfmt_java, additionally recognizing the .html extension with no need to put
982 <!--applet> to every applet file). You have to install the JDK and the
983 shell-script /usr/local/java/bin/javawrapper too. It works around the
984 brokenness of the Java filename handling. To add a Java binary, just create a
985 link to the class-file somewhere in the path.
987 2.3 /proc/sys/kernel - general kernel parameters
988 ------------------------------------------------
990 This directory reflects general kernel behaviors. As I've said before, the
991 contents depend on your configuration. Here you'll find the most important
992 files, along with descriptions of what they mean and how to use them.
997 The file contains three values; highwater, lowwater, and frequency.
999 It exists only when BSD-style process accounting is enabled. These values
1000 control its behavior. If the free space on the file system where the log lives
1001 goes below lowwater percentage, accounting suspends. If it goes above
1002 highwater percentage, accounting resumes. Frequency determines how often you
1003 check the amount of free space (value is in seconds). Default settings are: 4,
1004 2, and 30. That is, suspend accounting if there is less than 2 percent free;
1005 resume it if we have a value of 3 or more percent; consider information about
1006 the amount of free space valid for 30 seconds
1011 When the value in this file is 0, ctrl-alt-del is trapped and sent to the init
1012 program to handle a graceful restart. However, when the value is greater that
1013 zero, Linux's reaction to this key combination will be an immediate reboot,
1014 without syncing its dirty buffers.
1017 When a program (like dosemu) has the keyboard in raw mode, the
1018 ctrl-alt-del is intercepted by the program before it ever reaches the
1019 kernel tty layer, and it is up to the program to decide what to do with
1022 domainname and hostname
1023 -----------------------
1025 These files can be controlled to set the NIS domainname and hostname of your
1026 box. For the classic darkstar.frop.org a simple:
1028 # echo "darkstar" > /proc/sys/kernel/hostname
1029 # echo "frop.org" > /proc/sys/kernel/domainname
1032 would suffice to set your hostname and NIS domainname.
1034 osrelease, ostype and version
1035 -----------------------------
1037 The names make it pretty obvious what these fields contain:
1039 > cat /proc/sys/kernel/osrelease
1042 > cat /proc/sys/kernel/ostype
1045 > cat /proc/sys/kernel/version
1046 #4 Fri Oct 1 12:41:14 PDT 1999
1049 The files osrelease and ostype should be clear enough. Version needs a little
1050 more clarification. The #4 means that this is the 4th kernel built from this
1051 source base and the date after it indicates the time the kernel was built. The
1052 only way to tune these values is to rebuild the kernel.
1057 The value in this file represents the number of seconds the kernel waits
1058 before rebooting on a panic. When you use the software watchdog, the
1059 recommended setting is 60. If set to 0, the auto reboot after a kernel panic
1060 is disabled, which is the default setting.
1065 The four values in printk denote
1067 * default_message_loglevel,
1068 * minimum_console_loglevel and
1069 * default_console_loglevel
1072 These values influence printk() behavior when printing or logging error
1073 messages, which come from inside the kernel. See syslog(2) for more
1074 information on the different log levels.
1079 Messages with a higher priority than this will be printed to the console.
1081 default_message_level
1082 ---------------------
1084 Messages without an explicit priority will be printed with this priority.
1086 minimum_console_loglevel
1087 ------------------------
1089 Minimum (highest) value to which the console_loglevel can be set.
1091 default_console_loglevel
1092 ------------------------
1094 Default value for console_loglevel.
1099 This file shows the size of the generic SCSI (sg) buffer. At this point, you
1100 can't tune it yet, but you can change it at compile time by editing
1101 include/scsi/sg.h and changing the value of SG_BIG_BUFF.
1103 If you use a scanner with SANE (Scanner Access Now Easy) you might want to set
1104 this to a higher value. Refer to the SANE documentation on this issue.
1109 The location where the modprobe binary is located. The kernel uses this
1110 program to load modules on demand.
1112 2.4 /proc/sys/vm - The virtual memory subsystem
1113 -----------------------------------------------
1115 The files in this directory can be used to tune the operation of the virtual
1116 memory (VM) subsystem of the Linux kernel.
1121 Controls the tendency of the kernel to reclaim the memory which is used for
1122 caching of directory and inode objects.
1124 At the default value of vfs_cache_pressure=100 the kernel will attempt to
1125 reclaim dentries and inodes at a "fair" rate with respect to pagecache and
1126 swapcache reclaim. Decreasing vfs_cache_pressure causes the kernel to prefer
1127 to retain dentry and inode caches. Increasing vfs_cache_pressure beyond 100
1128 causes the kernel to prefer to reclaim dentries and inodes.
1130 dirty_background_ratio
1131 ----------------------
1133 Contains, as a percentage of total system memory, the number of pages at which
1134 the pdflush background writeback daemon will start writing out dirty data.
1139 Contains, as a percentage of total system memory, the number of pages at which
1140 a process which is generating disk writes will itself start writing out dirty
1143 dirty_writeback_centisecs
1144 -------------------------
1146 The pdflush writeback daemons will periodically wake up and write `old' data
1147 out to disk. This tunable expresses the interval between those wakeups, in
1148 100'ths of a second.
1150 Setting this to zero disables periodic writeback altogether.
1152 dirty_expire_centisecs
1153 ----------------------
1155 This tunable is used to define when dirty data is old enough to be eligible
1156 for writeout by the pdflush daemons. It is expressed in 100'ths of a second.
1157 Data which has been dirty in-memory for longer than this interval will be
1158 written out next time a pdflush daemon wakes up.
1160 lower_zone_protection
1161 ---------------------
1163 For some specialised workloads on highmem machines it is dangerous for
1164 the kernel to allow process memory to be allocated from the "lowmem"
1165 zone. This is because that memory could then be pinned via the mlock()
1166 system call, or by unavailability of swapspace.
1168 And on large highmem machines this lack of reclaimable lowmem memory
1171 So the Linux page allocator has a mechanism which prevents allocations
1172 which _could_ use highmem from using too much lowmem. This means that
1173 a certain amount of lowmem is defended from the possibility of being
1174 captured into pinned user memory.
1176 (The same argument applies to the old 16 megabyte ISA DMA region. This
1177 mechanism will also defend that region from allocations which could use
1180 The `lower_zone_protection' tunable determines how aggressive the kernel is
1181 in defending these lower zones. The default value is zero - no
1184 If you have a machine which uses highmem or ISA DMA and your
1185 applications are using mlock(), or if you are running with no swap then
1186 you probably should increase the lower_zone_protection setting.
1188 The units of this tunable are fairly vague. It is approximately equal
1189 to "megabytes". So setting lower_zone_protection=100 will protect around 100
1190 megabytes of the lowmem zone from user allocations. It will also make
1191 those 100 megabytes unavaliable for use by applications and by
1192 pagecache, so there is a cost.
1194 The effects of this tunable may be observed by monitoring
1195 /proc/meminfo:LowFree. Write a single huge file and observe the point
1196 at which LowFree ceases to fall.
1198 A reasonable value for lower_zone_protection is 100.
1203 page-cluster controls the number of pages which are written to swap in
1204 a single attempt. The swap I/O size.
1206 It is a logarithmic value - setting it to zero means "1 page", setting
1207 it to 1 means "2 pages", setting it to 2 means "4 pages", etc.
1209 The default value is three (eight pages at a time). There may be some
1210 small benefits in tuning this to a different value if your workload is
1216 This file contains one value. The following algorithm is used to decide if
1217 there's enough memory: if the value of overcommit_memory is positive, then
1218 there's always enough memory. This is a useful feature, since programs often
1219 malloc() huge amounts of memory 'just in case', while they only use a small
1220 part of it. Leaving this value at 0 will lead to the failure of such a huge
1221 malloc(), when in fact the system has enough memory for the program to run.
1223 On the other hand, enabling this feature can cause you to run out of memory
1224 and thrash the system to death, so large and/or important servers will want to
1225 set this value to 0.
1227 2.5 /proc/sys/dev - Device specific parameters
1228 ----------------------------------------------
1230 Currently there is only support for CDROM drives, and for those, there is only
1231 one read-only file containing information about the CD-ROM drives attached to
1234 >cat /proc/sys/dev/cdrom/info
1235 CD-ROM information, Id: cdrom.c 2.55 1999/04/25
1239 drive # of slots: 1 0
1243 Can change speed: 1 1
1244 Can select disk: 0 1
1245 Can read multisession: 1 1
1247 Reports media changed: 1 1
1251 You see two drives, sr0 and hdb, along with a list of their features.
1253 2.6 /proc/sys/sunrpc - Remote procedure calls
1254 ---------------------------------------------
1256 This directory contains four files, which enable or disable debugging for the
1257 RPC functions NFS, NFS-daemon, RPC and NLM. The default values are 0. They can
1258 be set to one to turn debugging on. (The default value is 0 for each)
1260 2.7 /proc/sys/net - Networking stuff
1261 ------------------------------------
1263 The interface to the networking parts of the kernel is located in
1264 /proc/sys/net. Table 2-3 shows all possible subdirectories. You may see only
1265 some of them, depending on your kernel's configuration.
1268 Table 2-3: Subdirectories in /proc/sys/net
1269 ..............................................................................
1270 Directory Content Directory Content
1271 core General parameter appletalk Appletalk protocol
1272 unix Unix domain sockets netrom NET/ROM
1273 802 E802 protocol ax25 AX25
1274 ethernet Ethernet protocol rose X.25 PLP layer
1275 ipv4 IP version 4 x25 X.25 protocol
1276 ipx IPX token-ring IBM token ring
1277 bridge Bridging decnet DEC net
1279 ..............................................................................
1281 We will concentrate on IP networking here. Since AX15, X.25, and DEC Net are
1282 only minor players in the Linux world, we'll skip them in this chapter. You'll
1283 find some short info on Appletalk and IPX further on in this chapter. Review
1284 the online documentation and the kernel source to get a detailed view of the
1285 parameters for those protocols. In this section we'll discuss the
1286 subdirectories printed in bold letters in the table above. As default values
1287 are suitable for most needs, there is no need to change these values.
1289 /proc/sys/net/core - Network core options
1290 -----------------------------------------
1295 The default setting of the socket receive buffer in bytes.
1300 The maximum receive socket buffer size in bytes.
1305 The default setting (in bytes) of the socket send buffer.
1310 The maximum send socket buffer size in bytes.
1312 message_burst and message_cost
1313 ------------------------------
1315 These parameters are used to limit the warning messages written to the kernel
1316 log from the networking code. They enforce a rate limit to make a
1317 denial-of-service attack impossible. A higher message_cost factor, results in
1318 fewer messages that will be written. Message_burst controls when messages will
1319 be dropped. The default settings limit warning messages to one every five
1325 Maximum number of packets, queued on the INPUT side, when the interface
1326 receives packets faster than kernel can process them.
1331 Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
1332 of struct cmsghdr structures with appended data.
1334 /proc/sys/net/unix - Parameters for Unix domain sockets
1335 -------------------------------------------------------
1337 There are only two files in this subdirectory. They control the delays for
1338 deleting and destroying socket descriptors.
1340 2.8 /proc/sys/net/ipv4 - IPV4 settings
1341 --------------------------------------
1343 IP version 4 is still the most used protocol in Unix networking. It will be
1344 replaced by IP version 6 in the next couple of years, but for the moment it's
1345 the de facto standard for the internet and is used in most networking
1346 environments around the world. Because of the importance of this protocol,
1347 we'll have a deeper look into the subtree controlling the behavior of the IPv4
1348 subsystem of the Linux kernel.
1350 Let's start with the entries in /proc/sys/net/ipv4.
1355 icmp_echo_ignore_all and icmp_echo_ignore_broadcasts
1356 ----------------------------------------------------
1358 Turn on (1) or off (0), if the kernel should ignore all ICMP ECHO requests, or
1359 just those to broadcast and multicast addresses.
1361 Please note that if you accept ICMP echo requests with a broadcast/multi\-cast
1362 destination address your network may be used as an exploder for denial of
1363 service packet flooding attacks to other hosts.
1365 icmp_destunreach_rate, icmp_echoreply_rate, icmp_paramprob_rate and icmp_timeexeed_rate
1366 ---------------------------------------------------------------------------------------
1368 Sets limits for sending ICMP packets to specific targets. A value of zero
1369 disables all limiting. Any positive value sets the maximum package rate in
1370 hundredth of a second (on Intel systems).
1378 This file contains the number one if the host received its IP configuration by
1379 RARP, BOOTP, DHCP or a similar mechanism. Otherwise it is zero.
1384 TTL (Time To Live) for IPv4 interfaces. This is simply the maximum number of
1385 hops a packet may travel.
1390 Enable dynamic socket address rewriting on interface address change. This is
1391 useful for dialup interface with changing IP addresses.
1396 Enable or disable forwarding of IP packages between interfaces. Changing this
1397 value resets all other parameters to their default values. They differ if the
1398 kernel is configured as host or router.
1403 Range of ports used by TCP and UDP to choose the local port. Contains two
1404 numbers, the first number is the lowest port, the second number the highest
1405 local port. Default is 1024-4999. Should be changed to 32768-61000 for
1411 Global switch to turn path MTU discovery off. It can also be set on a per
1412 socket basis by the applications or on a per route basis.
1417 Enable/disable debugging of IP masquerading.
1419 IP fragmentation settings
1420 -------------------------
1422 ipfrag_high_trash and ipfrag_low_trash
1423 --------------------------------------
1425 Maximum memory used to reassemble IP fragments. When ipfrag_high_thresh bytes
1426 of memory is allocated for this purpose, the fragment handler will toss
1427 packets until ipfrag_low_thresh is reached.
1432 Time in seconds to keep an IP fragment in memory.
1440 This file controls the use of the ECN bit in the IPv4 headers, this is a new
1441 feature about Explicit Congestion Notification, but some routers and firewalls
1442 block trafic that has this bit set, so it could be necessary to echo 0 to
1443 /proc/sys/net/ipv4/tcp_ecn, if you want to talk to this sites. For more info
1444 you could read RFC2481.
1446 tcp_retrans_collapse
1447 --------------------
1449 Bug-to-bug compatibility with some broken printers. On retransmit, try to send
1450 larger packets to work around bugs in certain TCP stacks. Can be turned off by
1453 tcp_keepalive_probes
1454 --------------------
1456 Number of keep alive probes TCP sends out, until it decides that the
1457 connection is broken.
1462 How often TCP sends out keep alive messages, when keep alive is enabled. The
1468 Number of times initial SYNs for a TCP connection attempt will be
1469 retransmitted. Should not be higher than 255. This is only the timeout for
1470 outgoing connections, for incoming connections the number of retransmits is
1471 defined by tcp_retries1.
1476 Enable select acknowledgments after RFC2018.
1481 Enable timestamps as defined in RFC1323.
1486 Enable the strict RFC793 interpretation of the TCP urgent pointer field. The
1487 default is to use the BSD compatible interpretation of the urgent pointer
1488 pointing to the first byte after the urgent data. The RFC793 interpretation is
1489 to have it point to the last byte of urgent data. Enabling this option may
1490 lead to interoperatibility problems. Disabled by default.
1495 Only valid when the kernel was compiled with CONFIG_SYNCOOKIES. Send out
1496 syncookies when the syn backlog queue of a socket overflows. This is to ward
1497 off the common 'syn flood attack'. Disabled by default.
1499 Note that the concept of a socket backlog is abandoned. This means the peer
1500 may not receive reliable error messages from an over loaded server with
1506 Enable window scaling as defined in RFC1323.
1511 The length of time in seconds it takes to receive a final FIN before the
1512 socket is always closed. This is strictly a violation of the TCP
1513 specification, but required to prevent denial-of-service attacks.
1518 Indicates how many keep alive probes are sent per slow timer run. Should not
1519 be set too high to prevent bursts.
1524 Length of the per socket backlog queue. Since Linux 2.2 the backlog specified
1525 in listen(2) only specifies the length of the backlog queue of already
1526 established sockets. When more connection requests arrive Linux starts to drop
1527 packets. When syncookies are enabled the packets are still answered and the
1528 maximum queue is effectively ignored.
1533 Defines how often an answer to a TCP connection request is retransmitted
1539 Defines how often a TCP packet is retransmitted before giving up.
1541 Interface specific settings
1542 ---------------------------
1544 In the directory /proc/sys/net/ipv4/conf you'll find one subdirectory for each
1545 interface the system knows about and one directory calls all. Changes in the
1546 all subdirectory affect all interfaces, whereas changes in the other
1547 subdirectories affect only one interface. All directories have the same
1553 This switch decides if the kernel accepts ICMP redirect messages or not. The
1554 default is 'yes' if the kernel is configured for a regular host and 'no' for a
1555 router configuration.
1560 Should source routed packages be accepted or declined. The default is
1561 dependent on the kernel configuration. It's 'yes' for routers and 'no' for
1567 Accept packets with source address 0.b.c.d with destinations not to this host
1568 as local ones. It is supposed that a BOOTP relay daemon will catch and forward
1571 The default is 0, since this feature is not implemented yet (kernel version
1577 Enable or disable IP forwarding on this interface.
1582 Log packets with source addresses with no known route to kernel log.
1587 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE and a
1588 multicast routing daemon is required.
1593 Does (1) or does not (0) perform proxy ARP.
1598 Integer value determines if a source validation should be made. 1 means yes, 0
1599 means no. Disabled by default, but local/broadcast address spoofing is always
1602 If you set this to 1 on a router that is the only connection for a network to
1603 the net, it will prevent spoofing attacks against your internal networks
1604 (external addresses can still be spoofed), without the need for additional
1610 Accept ICMP redirect messages only for gateways, listed in default gateway
1611 list. Enabled by default.
1616 If it is not set the kernel does not assume that different subnets on this
1617 device can communicate directly. Default setting is 'yes'.
1622 Determines whether to send ICMP redirects to other hosts.
1627 The directory /proc/sys/net/ipv4/route contains several file to control
1630 error_burst and error_cost
1631 --------------------------
1633 These parameters are used to limit how many ICMP destination unreachable to
1634 send from the host in question. ICMP destination unreachable messages are
1635 sent when we can not reach the next hop, while trying to transmit a packet.
1636 It will also print some error messages to kernel logs if someone is ignoring
1637 our ICMP redirects. The higher the error_cost factor is, the fewer
1638 destination unreachable and error messages will be let through. Error_burst
1639 controls when destination unreachable messages and error messages will be
1640 dropped. The default settings limit warning messages to five every second.
1645 Writing to this file results in a flush of the routing cache.
1647 gc_elasticity, gc_interval, gc_min_interval, gc_tresh, gc_timeout,
1648 gc_thresh, gc_thresh1, gc_thresh2, gc_thresh3
1649 --------------------------------------------------------------
1651 Values to control the frequency and behavior of the garbage collection
1652 algorithm for the routing cache.
1657 Maximum size of the routing cache. Old entries will be purged once the cache
1658 reached has this size.
1660 max_delay, min_delay
1661 --------------------
1663 Delays for flushing the routing cache.
1665 redirect_load, redirect_number
1666 ------------------------------
1668 Factors which determine if more ICPM redirects should be sent to a specific
1669 host. No redirects will be sent once the load limit or the maximum number of
1670 redirects has been reached.
1675 Timeout for redirects. After this period redirects will be sent again, even if
1676 this has been stopped, because the load or number limit has been reached.
1678 Network Neighbor handling
1679 -------------------------
1681 Settings about how to handle connections with direct neighbors (nodes attached
1682 to the same link) can be found in the directory /proc/sys/net/ipv4/neigh.
1684 As we saw it in the conf directory, there is a default subdirectory which
1685 holds the default values, and one directory for each interface. The contents
1686 of the directories are identical, with the single exception that the default
1687 settings contain additional options to set garbage collection parameters.
1689 In the interface directories you'll find the following entries:
1694 A base value used for computing the random reachable time value as specified
1700 The time, expressed in jiffies (1/100 sec), between retransmitted Neighbor
1701 Solicitation messages. Used for address resolution and to determine if a
1702 neighbor is unreachable.
1707 Maximum queue length for a pending arp request - the number of packets which
1708 are accepted from other layers while the ARP address is still resolved.
1713 Maximum for random delay of answers to neighbor solicitation messages in
1714 jiffies (1/100 sec). Not yet implemented (Linux does not have anycast support
1720 Maximum number of retries for unicast solicitation.
1725 Maximum number of retries for multicast solicitation.
1727 delay_first_probe_time
1728 ----------------------
1730 Delay for the first time probe if the neighbor is reachable. (see
1736 An ARP/neighbor entry is only replaced with a new one if the old is at least
1737 locktime old. This prevents ARP cache thrashing.
1742 Maximum time (real time is random [0..proxytime]) before answering to an ARP
1743 request for which we have an proxy ARP entry. In some cases, this is used to
1744 prevent network flooding.
1749 Maximum queue length of the delayed proxy arp timer. (see proxy_delay).
1754 Determines the number of requests to send to the user level ARP daemon. Use 0
1760 Determines how often to check for stale ARP entries. After an ARP entry is
1761 stale it will be resolved again (which is useful when an IP address migrates
1762 to another machine). When ucast_solicit is greater than 0 it first tries to
1763 send an ARP packet directly to the known host When that fails and
1764 mcast_solicit is greater than 0, an ARP request is broadcasted.
1769 The /proc/sys/net/appletalk directory holds the Appletalk configuration data
1770 when Appletalk is loaded. The configurable parameters are:
1775 The amount of time we keep an ARP entry before expiring it. Used to age out
1781 The amount of time we will spend trying to resolve an Appletalk address.
1783 aarp-retransmit-limit
1784 ---------------------
1786 The number of times we will retransmit a query before giving up.
1791 Controls the rate at which expires are checked.
1793 The directory /proc/net/appletalk holds the list of active Appletalk sockets
1796 The fields indicate the DDP type, the local address (in network:node format)
1797 the remote address, the size of the transmit pending queue, the size of the
1798 received queue (bytes waiting for applications to read) the state and the uid
1801 /proc/net/atalk_iface lists all the interfaces configured for appletalk.It
1802 shows the name of the interface, its Appletalk address, the network range on
1803 that address (or network number for phase 1 networks), and the status of the
1806 /proc/net/atalk_route lists each known network route. It lists the target
1807 (network) that the route leads to, the router (may be directly connected), the
1808 route flags, and the device the route is using.
1813 The IPX protocol has no tunable values in proc/sys/net.
1815 The IPX protocol does, however, provide proc/net/ipx. This lists each IPX
1816 socket giving the local and remote addresses in Novell format (that is
1817 network:node:port). In accordance with the strange Novell tradition,
1818 everything but the port is in hex. Not_Connected is displayed for sockets that
1819 are not tied to a specific remote address. The Tx and Rx queue sizes indicate
1820 the number of bytes pending for transmission and reception. The state
1821 indicates the state the socket is in and the uid is the owning uid of the
1824 The /proc/net/ipx_interface file lists all IPX interfaces. For each interface
1825 it gives the network number, the node number, and indicates if the network is
1826 the primary network. It also indicates which device it is bound to (or
1827 Internal for internal networks) and the Frame Type if appropriate. Linux
1828 supports 802.3, 802.2, 802.2 SNAP and DIX (Blue Book) ethernet framing for
1831 The /proc/net/ipx_route table holds a list of IPX routes. For each route it
1832 gives the destination network, the router node (or Directly) and the network
1833 address of the router (or Connected) for internal networks.
1835 2.11 /proc/sys/fs/mqueue - POSIX message queues filesystem
1836 ----------------------------------------------------------
1838 The "mqueue" filesystem provides the necessary kernel features to enable the
1839 creation of a user space library that implements the POSIX message queues
1840 API (as noted by the MSG tag in the POSIX 1003.1-2001 version of the System
1841 Interfaces specification.)
1843 The "mqueue" filesystem contains values for determining/setting the amount of
1844 resources used by the file system.
1846 /proc/sys/fs/mqueue/queues_max is a read/write file for setting/getting the
1847 maximum number of message queues allowed on the system.
1849 /proc/sys/fs/mqueue/msg_max is a read/write file for setting/getting the
1850 maximum number of messages in a queue value. In fact it is the limiting value
1851 for another (user) limit which is set in mq_open invocation. This attribute of
1852 a queue must be less or equal then msg_max.
1854 /proc/sys/fs/mqueue/msgsize_max is a read/write file for setting/getting the
1855 maximum message size value (it is every message queue's attribute set during
1859 ------------------------------------------------------------------------------
1861 ------------------------------------------------------------------------------
1862 Certain aspects of kernel behavior can be modified at runtime, without the
1863 need to recompile the kernel, or even to reboot the system. The files in the
1864 /proc/sys tree can not only be read, but also modified. You can use the echo
1865 command to write value into these files, thereby changing the default settings
1867 ------------------------------------------------------------------------------