1 Documentation for /proc/sys/net/*
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>
7 For general info and legal blurb, please look in README.
9 ==============================================================
11 This file contains the documentation for the sysctl files in
14 The interface to the networking parts of the kernel is located in
15 /proc/sys/net. The following table shows all possible subdirectories. You may
16 see only some of them, depending on your kernel's configuration.
19 Table : Subdirectories in /proc/sys/net
20 ..............................................................................
21 Directory Content Directory Content
22 core General parameter appletalk Appletalk protocol
23 unix Unix domain sockets netrom NET/ROM
24 802 E802 protocol ax25 AX25
25 ethernet Ethernet protocol rose X.25 PLP layer
26 ipv4 IP version 4 x25 X.25 protocol
27 ipx IPX token-ring IBM token ring
28 bridge Bridging decnet DEC net
29 ipv6 IP version 6 tipc TIPC
30 ..............................................................................
32 1. /proc/sys/net/core - Network core options
33 -------------------------------------------------------
38 This enables Berkeley Packet Filter Just in Time compiler.
39 Currently supported on x86_64 architecture, bpf_jit provides a framework
40 to speed packet filtering, the one used by tcpdump/libpcap for example.
42 0 - disable the JIT (default value)
44 2 - enable the JIT and ask the compiler to emit traces on kernel log.
49 The maximum number of packets that kernel can handle on a NAPI interrupt,
50 it's a Per-CPU variable.
56 The default queuing discipline to use for network devices. This allows
57 overriding the default of pfifo_fast with an alternative. Since the default
58 queuing discipline is created without additional parameters so is best suited
59 to queuing disciplines that work well without configuration like stochastic
60 fair queue (sfq), CoDel (codel) or fair queue CoDel (fq_codel). Don't use
61 queuing disciplines like Hierarchical Token Bucket or Deficit Round Robin
62 which require setting up classes and bandwidths. Note that physical multiqueue
63 interfaces still use mq as root qdisc, which in turn uses this default for its
64 leaves. Virtual devices (like e.g. lo or veth) ignore this setting and instead
70 Low latency busy poll timeout for socket reads. (needs CONFIG_NET_RX_BUSY_POLL)
71 Approximate time in us to busy loop waiting for packets on the device queue.
72 This sets the default value of the SO_BUSY_POLL socket option.
73 Can be set or overridden per socket by setting socket option SO_BUSY_POLL,
74 which is the preferred method of enabling. If you need to enable the feature
75 globally via sysctl, a value of 50 is recommended.
76 Will increase power usage.
81 Low latency busy poll timeout for poll and select. (needs CONFIG_NET_RX_BUSY_POLL)
82 Approximate time in us to busy loop waiting for events.
83 Recommended value depends on the number of sockets you poll on.
84 For several sockets 50, for several hundreds 100.
85 For more than that you probably want to use epoll.
86 Note that only sockets with SO_BUSY_POLL set will be busy polled,
87 so you want to either selectively set SO_BUSY_POLL on those sockets or set
88 sysctl.net.busy_read globally.
89 Will increase power usage.
95 The default setting of the socket receive buffer in bytes.
100 The maximum receive socket buffer size in bytes.
104 Allow processes to receive tx timestamps looped together with the original
105 packet contents. If disabled, transmit timestamp requests from unprivileged
106 processes are dropped unless socket option SOF_TIMESTAMPING_OPT_TSONLY is set.
113 The default setting (in bytes) of the socket send buffer.
118 The maximum send socket buffer size in bytes.
120 message_burst and message_cost
121 ------------------------------
123 These parameters are used to limit the warning messages written to the kernel
124 log from the networking code. They enforce a rate limit to make a
125 denial-of-service attack impossible. A higher message_cost factor, results in
126 fewer messages that will be written. Message_burst controls when messages will
127 be dropped. The default settings limit warning messages to one every five
133 This sysctl is now unused.
135 This was used to control console messages from the networking stack that
136 occur because of problems on the network like duplicate address or bad
139 These messages are now emitted at KERN_DEBUG and can generally be enabled
140 and controlled by the dynamic_debug facility.
145 Maximum number of packets taken from all interfaces in one polling cycle (NAPI
146 poll). In one polling cycle interfaces which are registered to polling are
147 probed in a round-robin manner.
152 Maximum number of packets, queued on the INPUT side, when the interface
153 receives packets faster than kernel can process them.
158 RSS (Receive Side Scaling) enabled drivers use a 40 bytes host key that is
160 Some user space might need to gather its content even if drivers do not
161 provide ethtool -x support yet.
163 myhost:~# cat /proc/sys/net/core/netdev_rss_key
164 84:50:f4:00:a8:15:d1:a7:e9:7f:1d:60:35:c7:47:25:42:97:74:ca:56:bb:b6:a1:d8: ... (52 bytes total)
166 File contains nul bytes if no driver ever called netdev_rss_key_fill() function.
168 /proc/sys/net/core/netdev_rss_key contains 52 bytes of key,
169 but most drivers only use 40 bytes of it.
171 myhost:~# ethtool -x eth0
172 RX flow hash indirection table for eth0 with 8 RX ring(s):
175 84:50:f4:00:a8:15:d1:a7:e9:7f:1d:60:35:c7:47:25:42:97:74:ca:56:bb:b6:a1:d8:43:e3:c9:0c:fd:17:55:c2:3a:4d:69:ed:f1:42:89
177 netdev_tstamp_prequeue
178 ----------------------
180 If set to 0, RX packet timestamps can be sampled after RPS processing, when
181 the target CPU processes packets. It might give some delay on timestamps, but
182 permit to distribute the load on several cpus.
184 If set to 1 (default), timestamps are sampled as soon as possible, before
190 Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
191 of struct cmsghdr structures with appended data.
193 2. /proc/sys/net/unix - Parameters for Unix domain sockets
194 -------------------------------------------------------
196 There is only one file in this directory.
197 unix_dgram_qlen limits the max number of datagrams queued in Unix domain
198 socket's buffer. It will not take effect unless PF_UNIX flag is specified.
201 3. /proc/sys/net/ipv4 - IPV4 settings
202 -------------------------------------------------------
203 Please see: Documentation/networking/ip-sysctl.txt and ipvs-sysctl.txt for
204 descriptions of these entries.
208 -------------------------------------------------------
210 The /proc/sys/net/appletalk directory holds the Appletalk configuration data
211 when Appletalk is loaded. The configurable parameters are:
216 The amount of time we keep an ARP entry before expiring it. Used to age out
222 The amount of time we will spend trying to resolve an Appletalk address.
224 aarp-retransmit-limit
225 ---------------------
227 The number of times we will retransmit a query before giving up.
232 Controls the rate at which expires are checked.
234 The directory /proc/net/appletalk holds the list of active Appletalk sockets
237 The fields indicate the DDP type, the local address (in network:node format)
238 the remote address, the size of the transmit pending queue, the size of the
239 received queue (bytes waiting for applications to read) the state and the uid
242 /proc/net/atalk_iface lists all the interfaces configured for appletalk.It
243 shows the name of the interface, its Appletalk address, the network range on
244 that address (or network number for phase 1 networks), and the status of the
247 /proc/net/atalk_route lists each known network route. It lists the target
248 (network) that the route leads to, the router (may be directly connected), the
249 route flags, and the device the route is using.
253 -------------------------------------------------------
255 The IPX protocol has no tunable values in proc/sys/net.
257 The IPX protocol does, however, provide proc/net/ipx. This lists each IPX
258 socket giving the local and remote addresses in Novell format (that is
259 network:node:port). In accordance with the strange Novell tradition,
260 everything but the port is in hex. Not_Connected is displayed for sockets that
261 are not tied to a specific remote address. The Tx and Rx queue sizes indicate
262 the number of bytes pending for transmission and reception. The state
263 indicates the state the socket is in and the uid is the owning uid of the
266 The /proc/net/ipx_interface file lists all IPX interfaces. For each interface
267 it gives the network number, the node number, and indicates if the network is
268 the primary network. It also indicates which device it is bound to (or
269 Internal for internal networks) and the Frame Type if appropriate. Linux
270 supports 802.3, 802.2, 802.2 SNAP and DIX (Blue Book) ethernet framing for
273 The /proc/net/ipx_route table holds a list of IPX routes. For each route it
274 gives the destination network, the router node (or Directly) and the network
275 address of the router (or Connected) for internal networks.
278 -------------------------------------------------------
283 The TIPC protocol now has a tunable for the receive memory, similar to the
284 tcp_rmem - i.e. a vector of 3 INTEGERs: (min, default, max)
286 # cat /proc/sys/net/tipc/tipc_rmem
287 4252725 34021800 68043600
290 The max value is set to CONN_OVERLOAD_LIMIT, and the default and min values
291 are scaled (shifted) versions of that same value. Note that the min value
292 is not at this point in time used in any meaningful way, but the triplet is
293 preserved in order to be consistent with things like tcp_rmem.
298 TIPC name table updates are distributed asynchronously in a cluster, without
299 any form of transaction handling. This means that different race scenarios are
300 possible. One such is that a name withdrawal sent out by one node and received
301 by another node may arrive after a second, overlapping name publication already
302 has been accepted from a third node, although the conflicting updates
303 originally may have been issued in the correct sequential order.
304 If named_timeout is nonzero, failed topology updates will be placed on a defer
305 queue until another event arrives that clears the error, or until the timeout
306 expires. Value is in milliseconds.