net.txt 7.0 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202
  1. Documentation for /proc/sys/net/* kernel version 2.4.0-test11-pre4
  2. (c) 1999 Terrehon Bowden <terrehon@pacbell.net>
  3. Bodo Bauer <bb@ricochet.net>
  4. (c) 2000 Jorge Nerin <comandante@zaralinux.com>
  5. (c) 2009 Shen Feng <shen@cn.fujitsu.com>
  6. For general info and legal blurb, please look in README.
  7. ==============================================================
  8. This file contains the documentation for the sysctl files in
  9. /proc/sys/net and is valid for Linux kernel version 2.4.0-test11-pre4.
  10. The interface to the networking parts of the kernel is located in
  11. /proc/sys/net. The following table shows all possible subdirectories.You may
  12. see only some of them, depending on your kernel's configuration.
  13. Table : Subdirectories in /proc/sys/net
  14. ..............................................................................
  15. Directory Content Directory Content
  16. core General parameter appletalk Appletalk protocol
  17. unix Unix domain sockets netrom NET/ROM
  18. 802 E802 protocol ax25 AX25
  19. ethernet Ethernet protocol rose X.25 PLP layer
  20. ipv4 IP version 4 x25 X.25 protocol
  21. ipx IPX token-ring IBM token ring
  22. bridge Bridging decnet DEC net
  23. ipv6 IP version 6
  24. ..............................................................................
  25. 1. /proc/sys/net/core - Network core options
  26. -------------------------------------------------------
  27. bpf_jit_enable
  28. --------------
  29. This enables Berkeley Packet Filter Just in Time compiler.
  30. Currently supported on x86_64 architecture, bpf_jit provides a framework
  31. to speed packet filtering, the one used by tcpdump/libpcap for example.
  32. Values :
  33. 0 - disable the JIT (default value)
  34. 1 - enable the JIT
  35. 2 - enable the JIT and ask the compiler to emit traces on kernel log.
  36. dev_weight
  37. --------------
  38. The maximum number of packets that kernel can handle on a NAPI interrupt,
  39. it's a Per-CPU variable.
  40. Default: 64
  41. rmem_default
  42. ------------
  43. The default setting of the socket receive buffer in bytes.
  44. rmem_max
  45. --------
  46. The maximum receive socket buffer size in bytes.
  47. wmem_default
  48. ------------
  49. The default setting (in bytes) of the socket send buffer.
  50. wmem_max
  51. --------
  52. The maximum send socket buffer size in bytes.
  53. message_burst and message_cost
  54. ------------------------------
  55. These parameters are used to limit the warning messages written to the kernel
  56. log from the networking code. They enforce a rate limit to make a
  57. denial-of-service attack impossible. A higher message_cost factor, results in
  58. fewer messages that will be written. Message_burst controls when messages will
  59. be dropped. The default settings limit warning messages to one every five
  60. seconds.
  61. warnings
  62. --------
  63. This controls console messages from the networking stack that can occur because
  64. of problems on the network like duplicate address or bad checksums. Normally,
  65. this should be enabled, but if the problem persists the messages can be
  66. disabled.
  67. netdev_budget
  68. -------------
  69. Maximum number of packets taken from all interfaces in one polling cycle (NAPI
  70. poll). In one polling cycle interfaces which are registered to polling are
  71. probed in a round-robin manner.
  72. netdev_max_backlog
  73. ------------------
  74. Maximum number of packets, queued on the INPUT side, when the interface
  75. receives packets faster than kernel can process them.
  76. netdev_tstamp_prequeue
  77. ----------------------
  78. If set to 0, RX packet timestamps can be sampled after RPS processing, when
  79. the target CPU processes packets. It might give some delay on timestamps, but
  80. permit to distribute the load on several cpus.
  81. If set to 1 (default), timestamps are sampled as soon as possible, before
  82. queueing.
  83. optmem_max
  84. ----------
  85. Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
  86. of struct cmsghdr structures with appended data.
  87. 2. /proc/sys/net/unix - Parameters for Unix domain sockets
  88. -------------------------------------------------------
  89. There is only one file in this directory.
  90. unix_dgram_qlen limits the max number of datagrams queued in Unix domain
  91. socket's buffer. It will not take effect unless PF_UNIX flag is specified.
  92. 3. /proc/sys/net/ipv4 - IPV4 settings
  93. -------------------------------------------------------
  94. Please see: Documentation/networking/ip-sysctl.txt and ipvs-sysctl.txt for
  95. descriptions of these entries.
  96. 4. Appletalk
  97. -------------------------------------------------------
  98. The /proc/sys/net/appletalk directory holds the Appletalk configuration data
  99. when Appletalk is loaded. The configurable parameters are:
  100. aarp-expiry-time
  101. ----------------
  102. The amount of time we keep an ARP entry before expiring it. Used to age out
  103. old hosts.
  104. aarp-resolve-time
  105. -----------------
  106. The amount of time we will spend trying to resolve an Appletalk address.
  107. aarp-retransmit-limit
  108. ---------------------
  109. The number of times we will retransmit a query before giving up.
  110. aarp-tick-time
  111. --------------
  112. Controls the rate at which expires are checked.
  113. The directory /proc/net/appletalk holds the list of active Appletalk sockets
  114. on a machine.
  115. The fields indicate the DDP type, the local address (in network:node format)
  116. the remote address, the size of the transmit pending queue, the size of the
  117. received queue (bytes waiting for applications to read) the state and the uid
  118. owning the socket.
  119. /proc/net/atalk_iface lists all the interfaces configured for appletalk.It
  120. shows the name of the interface, its Appletalk address, the network range on
  121. that address (or network number for phase 1 networks), and the status of the
  122. interface.
  123. /proc/net/atalk_route lists each known network route. It lists the target
  124. (network) that the route leads to, the router (may be directly connected), the
  125. route flags, and the device the route is using.
  126. 5. IPX
  127. -------------------------------------------------------
  128. The IPX protocol has no tunable values in proc/sys/net.
  129. The IPX protocol does, however, provide proc/net/ipx. This lists each IPX
  130. socket giving the local and remote addresses in Novell format (that is
  131. network:node:port). In accordance with the strange Novell tradition,
  132. everything but the port is in hex. Not_Connected is displayed for sockets that
  133. are not tied to a specific remote address. The Tx and Rx queue sizes indicate
  134. the number of bytes pending for transmission and reception. The state
  135. indicates the state the socket is in and the uid is the owning uid of the
  136. socket.
  137. The /proc/net/ipx_interface file lists all IPX interfaces. For each interface
  138. it gives the network number, the node number, and indicates if the network is
  139. the primary network. It also indicates which device it is bound to (or
  140. Internal for internal networks) and the Frame Type if appropriate. Linux
  141. supports 802.3, 802.2, 802.2 SNAP and DIX (Blue Book) ethernet framing for
  142. IPX.
  143. The /proc/net/ipx_route table holds a list of IPX routes. For each route it
  144. gives the destination network, the router node (or Directly) and the network
  145. address of the router (or Connected) for internal networks.