Loading...
1================================
2Documentation for /proc/sys/net/
3================================
4
5Copyright
6
7Copyright (c) 1999
8
9 - Terrehon Bowden <terrehon@pacbell.net>
10 - Bodo Bauer <bb@ricochet.net>
11
12Copyright (c) 2000
13
14 - Jorge Nerin <comandante@zaralinux.com>
15
16Copyright (c) 2009
17
18 - Shen Feng <shen@cn.fujitsu.com>
19
20For general info and legal blurb, please look in index.rst.
21
22------------------------------------------------------------------------------
23
24This file contains the documentation for the sysctl files in
25/proc/sys/net
26
27The interface to the networking parts of the kernel is located in
28/proc/sys/net. The following table shows all possible subdirectories. You may
29see only some of them, depending on your kernel's configuration.
30
31
32Table : Subdirectories in /proc/sys/net
33
34 ========= =================== = ========== ==================
35 Directory Content Directory Content
36 ========= =================== = ========== ==================
37 core General parameter appletalk Appletalk protocol
38 unix Unix domain sockets netrom NET/ROM
39 802 E802 protocol ax25 AX25
40 ethernet Ethernet protocol rose X.25 PLP layer
41 ipv4 IP version 4 x25 X.25 protocol
42 bridge Bridging decnet DEC net
43 ipv6 IP version 6 tipc TIPC
44 ========= =================== = ========== ==================
45
461. /proc/sys/net/core - Network core options
47============================================
48
49bpf_jit_enable
50--------------
51
52This enables the BPF Just in Time (JIT) compiler. BPF is a flexible
53and efficient infrastructure allowing to execute bytecode at various
54hook points. It is used in a number of Linux kernel subsystems such
55as networking (e.g. XDP, tc), tracing (e.g. kprobes, uprobes, tracepoints)
56and security (e.g. seccomp). LLVM has a BPF back end that can compile
57restricted C into a sequence of BPF instructions. After program load
58through bpf(2) and passing a verifier in the kernel, a JIT will then
59translate these BPF proglets into native CPU instructions. There are
60two flavors of JITs, the newer eBPF JIT currently supported on:
61
62 - x86_64
63 - x86_32
64 - arm64
65 - arm32
66 - ppc64
67 - sparc64
68 - mips64
69 - s390x
70 - riscv
71
72And the older cBPF JIT supported on the following archs:
73
74 - mips
75 - ppc
76 - sparc
77
78eBPF JITs are a superset of cBPF JITs, meaning the kernel will
79migrate cBPF instructions into eBPF instructions and then JIT
80compile them transparently. Older cBPF JITs can only translate
81tcpdump filters, seccomp rules, etc, but not mentioned eBPF
82programs loaded through bpf(2).
83
84Values:
85
86 - 0 - disable the JIT (default value)
87 - 1 - enable the JIT
88 - 2 - enable the JIT and ask the compiler to emit traces on kernel log.
89
90bpf_jit_harden
91--------------
92
93This enables hardening for the BPF JIT compiler. Supported are eBPF
94JIT backends. Enabling hardening trades off performance, but can
95mitigate JIT spraying.
96
97Values:
98
99 - 0 - disable JIT hardening (default value)
100 - 1 - enable JIT hardening for unprivileged users only
101 - 2 - enable JIT hardening for all users
102
103bpf_jit_kallsyms
104----------------
105
106When BPF JIT compiler is enabled, then compiled images are unknown
107addresses to the kernel, meaning they neither show up in traces nor
108in /proc/kallsyms. This enables export of these addresses, which can
109be used for debugging/tracing. If bpf_jit_harden is enabled, this
110feature is disabled.
111
112Values :
113
114 - 0 - disable JIT kallsyms export (default value)
115 - 1 - enable JIT kallsyms export for privileged users only
116
117bpf_jit_limit
118-------------
119
120This enforces a global limit for memory allocations to the BPF JIT
121compiler in order to reject unprivileged JIT requests once it has
122been surpassed. bpf_jit_limit contains the value of the global limit
123in bytes.
124
125dev_weight
126----------
127
128The maximum number of packets that kernel can handle on a NAPI interrupt,
129it's a Per-CPU variable. For drivers that support LRO or GRO_HW, a hardware
130aggregated packet is counted as one packet in this context.
131
132Default: 64
133
134dev_weight_rx_bias
135------------------
136
137RPS (e.g. RFS, aRFS) processing is competing with the registered NAPI poll function
138of the driver for the per softirq cycle netdev_budget. This parameter influences
139the proportion of the configured netdev_budget that is spent on RPS based packet
140processing during RX softirq cycles. It is further meant for making current
141dev_weight adaptable for asymmetric CPU needs on RX/TX side of the network stack.
142(see dev_weight_tx_bias) It is effective on a per CPU basis. Determination is based
143on dev_weight and is calculated multiplicative (dev_weight * dev_weight_rx_bias).
144
145Default: 1
146
147dev_weight_tx_bias
148------------------
149
150Scales the maximum number of packets that can be processed during a TX softirq cycle.
151Effective on a per CPU basis. Allows scaling of current dev_weight for asymmetric
152net stack processing needs. Be careful to avoid making TX softirq processing a CPU hog.
153
154Calculation is based on dev_weight (dev_weight * dev_weight_tx_bias).
155
156Default: 1
157
158default_qdisc
159-------------
160
161The default queuing discipline to use for network devices. This allows
162overriding the default of pfifo_fast with an alternative. Since the default
163queuing discipline is created without additional parameters so is best suited
164to queuing disciplines that work well without configuration like stochastic
165fair queue (sfq), CoDel (codel) or fair queue CoDel (fq_codel). Don't use
166queuing disciplines like Hierarchical Token Bucket or Deficit Round Robin
167which require setting up classes and bandwidths. Note that physical multiqueue
168interfaces still use mq as root qdisc, which in turn uses this default for its
169leaves. Virtual devices (like e.g. lo or veth) ignore this setting and instead
170default to noqueue.
171
172Default: pfifo_fast
173
174busy_read
175---------
176
177Low latency busy poll timeout for socket reads. (needs CONFIG_NET_RX_BUSY_POLL)
178Approximate time in us to busy loop waiting for packets on the device queue.
179This sets the default value of the SO_BUSY_POLL socket option.
180Can be set or overridden per socket by setting socket option SO_BUSY_POLL,
181which is the preferred method of enabling. If you need to enable the feature
182globally via sysctl, a value of 50 is recommended.
183
184Will increase power usage.
185
186Default: 0 (off)
187
188busy_poll
189----------------
190Low latency busy poll timeout for poll and select. (needs CONFIG_NET_RX_BUSY_POLL)
191Approximate time in us to busy loop waiting for events.
192Recommended value depends on the number of sockets you poll on.
193For several sockets 50, for several hundreds 100.
194For more than that you probably want to use epoll.
195Note that only sockets with SO_BUSY_POLL set will be busy polled,
196so you want to either selectively set SO_BUSY_POLL on those sockets or set
197sysctl.net.busy_read globally.
198
199Will increase power usage.
200
201Default: 0 (off)
202
203rmem_default
204------------
205
206The default setting of the socket receive buffer in bytes.
207
208rmem_max
209--------
210
211The maximum receive socket buffer size in bytes.
212
213tstamp_allow_data
214-----------------
215Allow processes to receive tx timestamps looped together with the original
216packet contents. If disabled, transmit timestamp requests from unprivileged
217processes are dropped unless socket option SOF_TIMESTAMPING_OPT_TSONLY is set.
218
219Default: 1 (on)
220
221
222wmem_default
223------------
224
225The default setting (in bytes) of the socket send buffer.
226
227wmem_max
228--------
229
230The maximum send socket buffer size in bytes.
231
232message_burst and message_cost
233------------------------------
234
235These parameters are used to limit the warning messages written to the kernel
236log from the networking code. They enforce a rate limit to make a
237denial-of-service attack impossible. A higher message_cost factor, results in
238fewer messages that will be written. Message_burst controls when messages will
239be dropped. The default settings limit warning messages to one every five
240seconds.
241
242warnings
243--------
244
245This sysctl is now unused.
246
247This was used to control console messages from the networking stack that
248occur because of problems on the network like duplicate address or bad
249checksums.
250
251These messages are now emitted at KERN_DEBUG and can generally be enabled
252and controlled by the dynamic_debug facility.
253
254netdev_budget
255-------------
256
257Maximum number of packets taken from all interfaces in one polling cycle (NAPI
258poll). In one polling cycle interfaces which are registered to polling are
259probed in a round-robin manner. Also, a polling cycle may not exceed
260netdev_budget_usecs microseconds, even if netdev_budget has not been
261exhausted.
262
263netdev_budget_usecs
264---------------------
265
266Maximum number of microseconds in one NAPI polling cycle. Polling
267will exit when either netdev_budget_usecs have elapsed during the
268poll cycle or the number of packets processed reaches netdev_budget.
269
270netdev_max_backlog
271------------------
272
273Maximum number of packets, queued on the INPUT side, when the interface
274receives packets faster than kernel can process them.
275
276netdev_rss_key
277--------------
278
279RSS (Receive Side Scaling) enabled drivers use a 40 bytes host key that is
280randomly generated.
281Some user space might need to gather its content even if drivers do not
282provide ethtool -x support yet.
283
284::
285
286 myhost:~# cat /proc/sys/net/core/netdev_rss_key
287 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)
288
289File contains nul bytes if no driver ever called netdev_rss_key_fill() function.
290
291Note:
292 /proc/sys/net/core/netdev_rss_key contains 52 bytes of key,
293 but most drivers only use 40 bytes of it.
294
295::
296
297 myhost:~# ethtool -x eth0
298 RX flow hash indirection table for eth0 with 8 RX ring(s):
299 0: 0 1 2 3 4 5 6 7
300 RSS hash key:
301 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
302
303netdev_tstamp_prequeue
304----------------------
305
306If set to 0, RX packet timestamps can be sampled after RPS processing, when
307the target CPU processes packets. It might give some delay on timestamps, but
308permit to distribute the load on several cpus.
309
310If set to 1 (default), timestamps are sampled as soon as possible, before
311queueing.
312
313optmem_max
314----------
315
316Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
317of struct cmsghdr structures with appended data.
318
319fb_tunnels_only_for_init_net
320----------------------------
321
322Controls if fallback tunnels (like tunl0, gre0, gretap0, erspan0,
323sit0, ip6tnl0, ip6gre0) are automatically created when a new
324network namespace is created, if corresponding tunnel is present
325in initial network namespace.
326If set to 1, these devices are not automatically created, and
327user space is responsible for creating them if needed.
328
329Default : 0 (for compatibility reasons)
330
331devconf_inherit_init_net
332------------------------
333
334Controls if a new network namespace should inherit all current
335settings under /proc/sys/net/{ipv4,ipv6}/conf/{all,default}/. By
336default, we keep the current behavior: for IPv4 we inherit all current
337settings from init_net and for IPv6 we reset all settings to default.
338
339If set to 1, both IPv4 and IPv6 settings are forced to inherit from
340current ones in init_net. If set to 2, both IPv4 and IPv6 settings are
341forced to reset to their default values.
342
343Default : 0 (for compatibility reasons)
344
3452. /proc/sys/net/unix - Parameters for Unix domain sockets
346----------------------------------------------------------
347
348There is only one file in this directory.
349unix_dgram_qlen limits the max number of datagrams queued in Unix domain
350socket's buffer. It will not take effect unless PF_UNIX flag is specified.
351
352
3533. /proc/sys/net/ipv4 - IPV4 settings
354-------------------------------------
355Please see: Documentation/networking/ip-sysctl.txt and ipvs-sysctl.txt for
356descriptions of these entries.
357
358
3594. Appletalk
360------------
361
362The /proc/sys/net/appletalk directory holds the Appletalk configuration data
363when Appletalk is loaded. The configurable parameters are:
364
365aarp-expiry-time
366----------------
367
368The amount of time we keep an ARP entry before expiring it. Used to age out
369old hosts.
370
371aarp-resolve-time
372-----------------
373
374The amount of time we will spend trying to resolve an Appletalk address.
375
376aarp-retransmit-limit
377---------------------
378
379The number of times we will retransmit a query before giving up.
380
381aarp-tick-time
382--------------
383
384Controls the rate at which expires are checked.
385
386The directory /proc/net/appletalk holds the list of active Appletalk sockets
387on a machine.
388
389The fields indicate the DDP type, the local address (in network:node format)
390the remote address, the size of the transmit pending queue, the size of the
391received queue (bytes waiting for applications to read) the state and the uid
392owning the socket.
393
394/proc/net/atalk_iface lists all the interfaces configured for appletalk.It
395shows the name of the interface, its Appletalk address, the network range on
396that address (or network number for phase 1 networks), and the status of the
397interface.
398
399/proc/net/atalk_route lists each known network route. It lists the target
400(network) that the route leads to, the router (may be directly connected), the
401route flags, and the device the route is using.
402
4035. TIPC
404-------
405
406tipc_rmem
407---------
408
409The TIPC protocol now has a tunable for the receive memory, similar to the
410tcp_rmem - i.e. a vector of 3 INTEGERs: (min, default, max)
411
412::
413
414 # cat /proc/sys/net/tipc/tipc_rmem
415 4252725 34021800 68043600
416 #
417
418The max value is set to CONN_OVERLOAD_LIMIT, and the default and min values
419are scaled (shifted) versions of that same value. Note that the min value
420is not at this point in time used in any meaningful way, but the triplet is
421preserved in order to be consistent with things like tcp_rmem.
422
423named_timeout
424-------------
425
426TIPC name table updates are distributed asynchronously in a cluster, without
427any form of transaction handling. This means that different race scenarios are
428possible. One such is that a name withdrawal sent out by one node and received
429by another node may arrive after a second, overlapping name publication already
430has been accepted from a third node, although the conflicting updates
431originally may have been issued in the correct sequential order.
432If named_timeout is nonzero, failed topology updates will be placed on a defer
433queue until another event arrives that clears the error, or until the timeout
434expires. Value is in milliseconds.
1================================
2Documentation for /proc/sys/net/
3================================
4
5Copyright
6
7Copyright (c) 1999
8
9 - Terrehon Bowden <terrehon@pacbell.net>
10 - Bodo Bauer <bb@ricochet.net>
11
12Copyright (c) 2000
13
14 - Jorge Nerin <comandante@zaralinux.com>
15
16Copyright (c) 2009
17
18 - Shen Feng <shen@cn.fujitsu.com>
19
20For general info and legal blurb, please look in index.rst.
21
22------------------------------------------------------------------------------
23
24This file contains the documentation for the sysctl files in
25/proc/sys/net
26
27The interface to the networking parts of the kernel is located in
28/proc/sys/net. The following table shows all possible subdirectories. You may
29see only some of them, depending on your kernel's configuration.
30
31
32Table : Subdirectories in /proc/sys/net
33
34 ========= =================== = ========== ==================
35 Directory Content Directory Content
36 ========= =================== = ========== ==================
37 core General parameter appletalk Appletalk protocol
38 unix Unix domain sockets netrom NET/ROM
39 802 E802 protocol ax25 AX25
40 ethernet Ethernet protocol rose X.25 PLP layer
41 ipv4 IP version 4 x25 X.25 protocol
42 bridge Bridging decnet DEC net
43 ipv6 IP version 6 tipc TIPC
44 ========= =================== = ========== ==================
45
461. /proc/sys/net/core - Network core options
47============================================
48
49bpf_jit_enable
50--------------
51
52This enables the BPF Just in Time (JIT) compiler. BPF is a flexible
53and efficient infrastructure allowing to execute bytecode at various
54hook points. It is used in a number of Linux kernel subsystems such
55as networking (e.g. XDP, tc), tracing (e.g. kprobes, uprobes, tracepoints)
56and security (e.g. seccomp). LLVM has a BPF back end that can compile
57restricted C into a sequence of BPF instructions. After program load
58through bpf(2) and passing a verifier in the kernel, a JIT will then
59translate these BPF proglets into native CPU instructions. There are
60two flavors of JITs, the newer eBPF JIT currently supported on:
61
62 - x86_64
63 - x86_32
64 - arm64
65 - arm32
66 - ppc64
67 - sparc64
68 - mips64
69 - s390x
70 - riscv64
71 - riscv32
72
73And the older cBPF JIT supported on the following archs:
74
75 - mips
76 - ppc
77 - sparc
78
79eBPF JITs are a superset of cBPF JITs, meaning the kernel will
80migrate cBPF instructions into eBPF instructions and then JIT
81compile them transparently. Older cBPF JITs can only translate
82tcpdump filters, seccomp rules, etc, but not mentioned eBPF
83programs loaded through bpf(2).
84
85Values:
86
87 - 0 - disable the JIT (default value)
88 - 1 - enable the JIT
89 - 2 - enable the JIT and ask the compiler to emit traces on kernel log.
90
91bpf_jit_harden
92--------------
93
94This enables hardening for the BPF JIT compiler. Supported are eBPF
95JIT backends. Enabling hardening trades off performance, but can
96mitigate JIT spraying.
97
98Values:
99
100 - 0 - disable JIT hardening (default value)
101 - 1 - enable JIT hardening for unprivileged users only
102 - 2 - enable JIT hardening for all users
103
104bpf_jit_kallsyms
105----------------
106
107When BPF JIT compiler is enabled, then compiled images are unknown
108addresses to the kernel, meaning they neither show up in traces nor
109in /proc/kallsyms. This enables export of these addresses, which can
110be used for debugging/tracing. If bpf_jit_harden is enabled, this
111feature is disabled.
112
113Values :
114
115 - 0 - disable JIT kallsyms export (default value)
116 - 1 - enable JIT kallsyms export for privileged users only
117
118bpf_jit_limit
119-------------
120
121This enforces a global limit for memory allocations to the BPF JIT
122compiler in order to reject unprivileged JIT requests once it has
123been surpassed. bpf_jit_limit contains the value of the global limit
124in bytes.
125
126dev_weight
127----------
128
129The maximum number of packets that kernel can handle on a NAPI interrupt,
130it's a Per-CPU variable. For drivers that support LRO or GRO_HW, a hardware
131aggregated packet is counted as one packet in this context.
132
133Default: 64
134
135dev_weight_rx_bias
136------------------
137
138RPS (e.g. RFS, aRFS) processing is competing with the registered NAPI poll function
139of the driver for the per softirq cycle netdev_budget. This parameter influences
140the proportion of the configured netdev_budget that is spent on RPS based packet
141processing during RX softirq cycles. It is further meant for making current
142dev_weight adaptable for asymmetric CPU needs on RX/TX side of the network stack.
143(see dev_weight_tx_bias) It is effective on a per CPU basis. Determination is based
144on dev_weight and is calculated multiplicative (dev_weight * dev_weight_rx_bias).
145
146Default: 1
147
148dev_weight_tx_bias
149------------------
150
151Scales the maximum number of packets that can be processed during a TX softirq cycle.
152Effective on a per CPU basis. Allows scaling of current dev_weight for asymmetric
153net stack processing needs. Be careful to avoid making TX softirq processing a CPU hog.
154
155Calculation is based on dev_weight (dev_weight * dev_weight_tx_bias).
156
157Default: 1
158
159default_qdisc
160-------------
161
162The default queuing discipline to use for network devices. This allows
163overriding the default of pfifo_fast with an alternative. Since the default
164queuing discipline is created without additional parameters so is best suited
165to queuing disciplines that work well without configuration like stochastic
166fair queue (sfq), CoDel (codel) or fair queue CoDel (fq_codel). Don't use
167queuing disciplines like Hierarchical Token Bucket or Deficit Round Robin
168which require setting up classes and bandwidths. Note that physical multiqueue
169interfaces still use mq as root qdisc, which in turn uses this default for its
170leaves. Virtual devices (like e.g. lo or veth) ignore this setting and instead
171default to noqueue.
172
173Default: pfifo_fast
174
175busy_read
176---------
177
178Low latency busy poll timeout for socket reads. (needs CONFIG_NET_RX_BUSY_POLL)
179Approximate time in us to busy loop waiting for packets on the device queue.
180This sets the default value of the SO_BUSY_POLL socket option.
181Can be set or overridden per socket by setting socket option SO_BUSY_POLL,
182which is the preferred method of enabling. If you need to enable the feature
183globally via sysctl, a value of 50 is recommended.
184
185Will increase power usage.
186
187Default: 0 (off)
188
189busy_poll
190----------------
191Low latency busy poll timeout for poll and select. (needs CONFIG_NET_RX_BUSY_POLL)
192Approximate time in us to busy loop waiting for events.
193Recommended value depends on the number of sockets you poll on.
194For several sockets 50, for several hundreds 100.
195For more than that you probably want to use epoll.
196Note that only sockets with SO_BUSY_POLL set will be busy polled,
197so you want to either selectively set SO_BUSY_POLL on those sockets or set
198sysctl.net.busy_read globally.
199
200Will increase power usage.
201
202Default: 0 (off)
203
204rmem_default
205------------
206
207The default setting of the socket receive buffer in bytes.
208
209rmem_max
210--------
211
212The maximum receive socket buffer size in bytes.
213
214tstamp_allow_data
215-----------------
216Allow processes to receive tx timestamps looped together with the original
217packet contents. If disabled, transmit timestamp requests from unprivileged
218processes are dropped unless socket option SOF_TIMESTAMPING_OPT_TSONLY is set.
219
220Default: 1 (on)
221
222
223wmem_default
224------------
225
226The default setting (in bytes) of the socket send buffer.
227
228wmem_max
229--------
230
231The maximum send socket buffer size in bytes.
232
233message_burst and message_cost
234------------------------------
235
236These parameters are used to limit the warning messages written to the kernel
237log from the networking code. They enforce a rate limit to make a
238denial-of-service attack impossible. A higher message_cost factor, results in
239fewer messages that will be written. Message_burst controls when messages will
240be dropped. The default settings limit warning messages to one every five
241seconds.
242
243warnings
244--------
245
246This sysctl is now unused.
247
248This was used to control console messages from the networking stack that
249occur because of problems on the network like duplicate address or bad
250checksums.
251
252These messages are now emitted at KERN_DEBUG and can generally be enabled
253and controlled by the dynamic_debug facility.
254
255netdev_budget
256-------------
257
258Maximum number of packets taken from all interfaces in one polling cycle (NAPI
259poll). In one polling cycle interfaces which are registered to polling are
260probed in a round-robin manner. Also, a polling cycle may not exceed
261netdev_budget_usecs microseconds, even if netdev_budget has not been
262exhausted.
263
264netdev_budget_usecs
265---------------------
266
267Maximum number of microseconds in one NAPI polling cycle. Polling
268will exit when either netdev_budget_usecs have elapsed during the
269poll cycle or the number of packets processed reaches netdev_budget.
270
271netdev_max_backlog
272------------------
273
274Maximum number of packets, queued on the INPUT side, when the interface
275receives packets faster than kernel can process them.
276
277netdev_rss_key
278--------------
279
280RSS (Receive Side Scaling) enabled drivers use a 40 bytes host key that is
281randomly generated.
282Some user space might need to gather its content even if drivers do not
283provide ethtool -x support yet.
284
285::
286
287 myhost:~# cat /proc/sys/net/core/netdev_rss_key
288 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)
289
290File contains nul bytes if no driver ever called netdev_rss_key_fill() function.
291
292Note:
293 /proc/sys/net/core/netdev_rss_key contains 52 bytes of key,
294 but most drivers only use 40 bytes of it.
295
296::
297
298 myhost:~# ethtool -x eth0
299 RX flow hash indirection table for eth0 with 8 RX ring(s):
300 0: 0 1 2 3 4 5 6 7
301 RSS hash key:
302 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
303
304netdev_tstamp_prequeue
305----------------------
306
307If set to 0, RX packet timestamps can be sampled after RPS processing, when
308the target CPU processes packets. It might give some delay on timestamps, but
309permit to distribute the load on several cpus.
310
311If set to 1 (default), timestamps are sampled as soon as possible, before
312queueing.
313
314optmem_max
315----------
316
317Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
318of struct cmsghdr structures with appended data.
319
320fb_tunnels_only_for_init_net
321----------------------------
322
323Controls if fallback tunnels (like tunl0, gre0, gretap0, erspan0,
324sit0, ip6tnl0, ip6gre0) are automatically created when a new
325network namespace is created, if corresponding tunnel is present
326in initial network namespace.
327If set to 1, these devices are not automatically created, and
328user space is responsible for creating them if needed.
329
330Default : 0 (for compatibility reasons)
331
332devconf_inherit_init_net
333------------------------
334
335Controls if a new network namespace should inherit all current
336settings under /proc/sys/net/{ipv4,ipv6}/conf/{all,default}/. By
337default, we keep the current behavior: for IPv4 we inherit all current
338settings from init_net and for IPv6 we reset all settings to default.
339
340If set to 1, both IPv4 and IPv6 settings are forced to inherit from
341current ones in init_net. If set to 2, both IPv4 and IPv6 settings are
342forced to reset to their default values. If set to 3, both IPv4 and IPv6
343settings are forced to inherit from current ones in the netns where this
344new netns has been created.
345
346Default : 0 (for compatibility reasons)
347
3482. /proc/sys/net/unix - Parameters for Unix domain sockets
349----------------------------------------------------------
350
351There is only one file in this directory.
352unix_dgram_qlen limits the max number of datagrams queued in Unix domain
353socket's buffer. It will not take effect unless PF_UNIX flag is specified.
354
355
3563. /proc/sys/net/ipv4 - IPV4 settings
357-------------------------------------
358Please see: Documentation/networking/ip-sysctl.rst and
359Documentation/admin-guide/sysctl/net.rst for descriptions of these entries.
360
361
3624. Appletalk
363------------
364
365The /proc/sys/net/appletalk directory holds the Appletalk configuration data
366when Appletalk is loaded. The configurable parameters are:
367
368aarp-expiry-time
369----------------
370
371The amount of time we keep an ARP entry before expiring it. Used to age out
372old hosts.
373
374aarp-resolve-time
375-----------------
376
377The amount of time we will spend trying to resolve an Appletalk address.
378
379aarp-retransmit-limit
380---------------------
381
382The number of times we will retransmit a query before giving up.
383
384aarp-tick-time
385--------------
386
387Controls the rate at which expires are checked.
388
389The directory /proc/net/appletalk holds the list of active Appletalk sockets
390on a machine.
391
392The fields indicate the DDP type, the local address (in network:node format)
393the remote address, the size of the transmit pending queue, the size of the
394received queue (bytes waiting for applications to read) the state and the uid
395owning the socket.
396
397/proc/net/atalk_iface lists all the interfaces configured for appletalk.It
398shows the name of the interface, its Appletalk address, the network range on
399that address (or network number for phase 1 networks), and the status of the
400interface.
401
402/proc/net/atalk_route lists each known network route. It lists the target
403(network) that the route leads to, the router (may be directly connected), the
404route flags, and the device the route is using.
405
4065. TIPC
407-------
408
409tipc_rmem
410---------
411
412The TIPC protocol now has a tunable for the receive memory, similar to the
413tcp_rmem - i.e. a vector of 3 INTEGERs: (min, default, max)
414
415::
416
417 # cat /proc/sys/net/tipc/tipc_rmem
418 4252725 34021800 68043600
419 #
420
421The max value is set to CONN_OVERLOAD_LIMIT, and the default and min values
422are scaled (shifted) versions of that same value. Note that the min value
423is not at this point in time used in any meaningful way, but the triplet is
424preserved in order to be consistent with things like tcp_rmem.
425
426named_timeout
427-------------
428
429TIPC name table updates are distributed asynchronously in a cluster, without
430any form of transaction handling. This means that different race scenarios are
431possible. One such is that a name withdrawal sent out by one node and received
432by another node may arrive after a second, overlapping name publication already
433has been accepted from a third node, although the conflicting updates
434originally may have been issued in the correct sequential order.
435If named_timeout is nonzero, failed topology updates will be placed on a defer
436queue until another event arrives that clears the error, or until the timeout
437expires. Value is in milliseconds.