xref: /linux/Documentation/admin-guide/sysctl/net.rst (revision 3652117f854819a148ff0fbe4492587d3520b5e5)
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 ========= =================== = ========== ===================
37 802       E802 protocol         mptcp      Multipath TCP
38 appletalk Appletalk protocol    netfilter  Network Filter
39 ax25      AX25                  netrom     NET/ROM
40 bridge    Bridging              rose       X.25 PLP layer
41 core      General parameter     tipc       TIPC
42 ethernet  Ethernet protocol     unix       Unix domain sockets
43 ipv4      IP version 4          x25        X.25 protocol
44 ipv6      IP version 6
45 ========= =================== = ========== ===================
46
471. /proc/sys/net/core - Network core options
48============================================
49
50bpf_jit_enable
51--------------
52
53This enables the BPF Just in Time (JIT) compiler. BPF is a flexible
54and efficient infrastructure allowing to execute bytecode at various
55hook points. It is used in a number of Linux kernel subsystems such
56as networking (e.g. XDP, tc), tracing (e.g. kprobes, uprobes, tracepoints)
57and security (e.g. seccomp). LLVM has a BPF back end that can compile
58restricted C into a sequence of BPF instructions. After program load
59through bpf(2) and passing a verifier in the kernel, a JIT will then
60translate these BPF proglets into native CPU instructions. There are
61two flavors of JITs, the newer eBPF JIT currently supported on:
62
63  - x86_64
64  - x86_32
65  - arm64
66  - arm32
67  - ppc64
68  - ppc32
69  - sparc64
70  - mips64
71  - s390x
72  - riscv64
73  - riscv32
74  - loongarch64
75
76And the older cBPF JIT supported on the following archs:
77
78  - mips
79  - sparc
80
81eBPF JITs are a superset of cBPF JITs, meaning the kernel will
82migrate cBPF instructions into eBPF instructions and then JIT
83compile them transparently. Older cBPF JITs can only translate
84tcpdump filters, seccomp rules, etc, but not mentioned eBPF
85programs loaded through bpf(2).
86
87Values:
88
89	- 0 - disable the JIT (default value)
90	- 1 - enable the JIT
91	- 2 - enable the JIT and ask the compiler to emit traces on kernel log.
92
93bpf_jit_harden
94--------------
95
96This enables hardening for the BPF JIT compiler. Supported are eBPF
97JIT backends. Enabling hardening trades off performance, but can
98mitigate JIT spraying.
99
100Values:
101
102	- 0 - disable JIT hardening (default value)
103	- 1 - enable JIT hardening for unprivileged users only
104	- 2 - enable JIT hardening for all users
105
106where "privileged user" in this context means a process having
107CAP_BPF or CAP_SYS_ADMIN in the root user name space.
108
109bpf_jit_kallsyms
110----------------
111
112When BPF JIT compiler is enabled, then compiled images are unknown
113addresses to the kernel, meaning they neither show up in traces nor
114in /proc/kallsyms. This enables export of these addresses, which can
115be used for debugging/tracing. If bpf_jit_harden is enabled, this
116feature is disabled.
117
118Values :
119
120	- 0 - disable JIT kallsyms export (default value)
121	- 1 - enable JIT kallsyms export for privileged users only
122
123bpf_jit_limit
124-------------
125
126This enforces a global limit for memory allocations to the BPF JIT
127compiler in order to reject unprivileged JIT requests once it has
128been surpassed. bpf_jit_limit contains the value of the global limit
129in bytes.
130
131dev_weight
132----------
133
134The maximum number of packets that kernel can handle on a NAPI interrupt,
135it's a Per-CPU variable. For drivers that support LRO or GRO_HW, a hardware
136aggregated packet is counted as one packet in this context.
137
138Default: 64
139
140dev_weight_rx_bias
141------------------
142
143RPS (e.g. RFS, aRFS) processing is competing with the registered NAPI poll function
144of the driver for the per softirq cycle netdev_budget. This parameter influences
145the proportion of the configured netdev_budget that is spent on RPS based packet
146processing during RX softirq cycles. It is further meant for making current
147dev_weight adaptable for asymmetric CPU needs on RX/TX side of the network stack.
148(see dev_weight_tx_bias) It is effective on a per CPU basis. Determination is based
149on dev_weight and is calculated multiplicative (dev_weight * dev_weight_rx_bias).
150
151Default: 1
152
153dev_weight_tx_bias
154------------------
155
156Scales the maximum number of packets that can be processed during a TX softirq cycle.
157Effective on a per CPU basis. Allows scaling of current dev_weight for asymmetric
158net stack processing needs. Be careful to avoid making TX softirq processing a CPU hog.
159
160Calculation is based on dev_weight (dev_weight * dev_weight_tx_bias).
161
162Default: 1
163
164default_qdisc
165-------------
166
167The default queuing discipline to use for network devices. This allows
168overriding the default of pfifo_fast with an alternative. Since the default
169queuing discipline is created without additional parameters so is best suited
170to queuing disciplines that work well without configuration like stochastic
171fair queue (sfq), CoDel (codel) or fair queue CoDel (fq_codel). Don't use
172queuing disciplines like Hierarchical Token Bucket or Deficit Round Robin
173which require setting up classes and bandwidths. Note that physical multiqueue
174interfaces still use mq as root qdisc, which in turn uses this default for its
175leaves. Virtual devices (like e.g. lo or veth) ignore this setting and instead
176default to noqueue.
177
178Default: pfifo_fast
179
180busy_read
181---------
182
183Low latency busy poll timeout for socket reads. (needs CONFIG_NET_RX_BUSY_POLL)
184Approximate time in us to busy loop waiting for packets on the device queue.
185This sets the default value of the SO_BUSY_POLL socket option.
186Can be set or overridden per socket by setting socket option SO_BUSY_POLL,
187which is the preferred method of enabling. If you need to enable the feature
188globally via sysctl, a value of 50 is recommended.
189
190Will increase power usage.
191
192Default: 0 (off)
193
194busy_poll
195----------------
196Low latency busy poll timeout for poll and select. (needs CONFIG_NET_RX_BUSY_POLL)
197Approximate time in us to busy loop waiting for events.
198Recommended value depends on the number of sockets you poll on.
199For several sockets 50, for several hundreds 100.
200For more than that you probably want to use epoll.
201Note that only sockets with SO_BUSY_POLL set will be busy polled,
202so you want to either selectively set SO_BUSY_POLL on those sockets or set
203sysctl.net.busy_read globally.
204
205Will increase power usage.
206
207Default: 0 (off)
208
209rmem_default
210------------
211
212The default setting of the socket receive buffer in bytes.
213
214rmem_max
215--------
216
217The maximum receive socket buffer size in bytes.
218
219rps_default_mask
220----------------
221
222The default RPS CPU mask used on newly created network devices. An empty
223mask means RPS disabled by default.
224
225tstamp_allow_data
226-----------------
227Allow processes to receive tx timestamps looped together with the original
228packet contents. If disabled, transmit timestamp requests from unprivileged
229processes are dropped unless socket option SOF_TIMESTAMPING_OPT_TSONLY is set.
230
231Default: 1 (on)
232
233
234wmem_default
235------------
236
237The default setting (in bytes) of the socket send buffer.
238
239wmem_max
240--------
241
242The maximum send socket buffer size in bytes.
243
244message_burst and message_cost
245------------------------------
246
247These parameters  are used to limit the warning messages written to the kernel
248log from  the  networking  code.  They  enforce  a  rate  limit  to  make  a
249denial-of-service attack  impossible. A higher message_cost factor, results in
250fewer messages that will be written. Message_burst controls when messages will
251be dropped.  The  default  settings  limit  warning messages to one every five
252seconds.
253
254warnings
255--------
256
257This sysctl is now unused.
258
259This was used to control console messages from the networking stack that
260occur because of problems on the network like duplicate address or bad
261checksums.
262
263These messages are now emitted at KERN_DEBUG and can generally be enabled
264and controlled by the dynamic_debug facility.
265
266netdev_budget
267-------------
268
269Maximum number of packets taken from all interfaces in one polling cycle (NAPI
270poll). In one polling cycle interfaces which are registered to polling are
271probed in a round-robin manner. Also, a polling cycle may not exceed
272netdev_budget_usecs microseconds, even if netdev_budget has not been
273exhausted.
274
275netdev_budget_usecs
276---------------------
277
278Maximum number of microseconds in one NAPI polling cycle. Polling
279will exit when either netdev_budget_usecs have elapsed during the
280poll cycle or the number of packets processed reaches netdev_budget.
281
282netdev_max_backlog
283------------------
284
285Maximum number of packets, queued on the INPUT side, when the interface
286receives packets faster than kernel can process them.
287
288netdev_rss_key
289--------------
290
291RSS (Receive Side Scaling) enabled drivers use a 40 bytes host key that is
292randomly generated.
293Some user space might need to gather its content even if drivers do not
294provide ethtool -x support yet.
295
296::
297
298  myhost:~# cat /proc/sys/net/core/netdev_rss_key
299  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)
300
301File contains nul bytes if no driver ever called netdev_rss_key_fill() function.
302
303Note:
304  /proc/sys/net/core/netdev_rss_key contains 52 bytes of key,
305  but most drivers only use 40 bytes of it.
306
307::
308
309  myhost:~# ethtool -x eth0
310  RX flow hash indirection table for eth0 with 8 RX ring(s):
311      0:    0     1     2     3     4     5     6     7
312  RSS hash key:
313  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
314
315netdev_tstamp_prequeue
316----------------------
317
318If set to 0, RX packet timestamps can be sampled after RPS processing, when
319the target CPU processes packets. It might give some delay on timestamps, but
320permit to distribute the load on several cpus.
321
322If set to 1 (default), timestamps are sampled as soon as possible, before
323queueing.
324
325netdev_unregister_timeout_secs
326------------------------------
327
328Unregister network device timeout in seconds.
329This option controls the timeout (in seconds) used to issue a warning while
330waiting for a network device refcount to drop to 0 during device
331unregistration. A lower value may be useful during bisection to detect
332a leaked reference faster. A larger value may be useful to prevent false
333warnings on slow/loaded systems.
334Default value is 10, minimum 1, maximum 3600.
335
336skb_defer_max
337-------------
338
339Max size (in skbs) of the per-cpu list of skbs being freed
340by the cpu which allocated them. Used by TCP stack so far.
341
342Default: 64
343
344optmem_max
345----------
346
347Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence
348of struct cmsghdr structures with appended data.
349
350fb_tunnels_only_for_init_net
351----------------------------
352
353Controls if fallback tunnels (like tunl0, gre0, gretap0, erspan0,
354sit0, ip6tnl0, ip6gre0) are automatically created. There are 3 possibilities
355(a) value = 0; respective fallback tunnels are created when module is
356loaded in every net namespaces (backward compatible behavior).
357(b) value = 1; [kcmd value: initns] respective fallback tunnels are
358created only in init net namespace and every other net namespace will
359not have them.
360(c) value = 2; [kcmd value: none] fallback tunnels are not created
361when a module is loaded in any of the net namespace. Setting value to
362"2" is pointless after boot if these modules are built-in, so there is
363a kernel command-line option that can change this default. Please refer to
364Documentation/admin-guide/kernel-parameters.txt for additional details.
365
366Not creating fallback tunnels gives control to userspace to create
367whatever is needed only and avoid creating devices which are redundant.
368
369Default : 0  (for compatibility reasons)
370
371devconf_inherit_init_net
372------------------------
373
374Controls if a new network namespace should inherit all current
375settings under /proc/sys/net/{ipv4,ipv6}/conf/{all,default}/. By
376default, we keep the current behavior: for IPv4 we inherit all current
377settings from init_net and for IPv6 we reset all settings to default.
378
379If set to 1, both IPv4 and IPv6 settings are forced to inherit from
380current ones in init_net. If set to 2, both IPv4 and IPv6 settings are
381forced to reset to their default values. If set to 3, both IPv4 and IPv6
382settings are forced to inherit from current ones in the netns where this
383new netns has been created.
384
385Default : 0  (for compatibility reasons)
386
387txrehash
388--------
389
390Controls default hash rethink behaviour on socket when SO_TXREHASH option is set
391to SOCK_TXREHASH_DEFAULT (i. e. not overridden by setsockopt).
392
393If set to 1 (default), hash rethink is performed on listening socket.
394If set to 0, hash rethink is not performed.
395
396gro_normal_batch
397----------------
398
399Maximum number of the segments to batch up on output of GRO. When a packet
400exits GRO, either as a coalesced superframe or as an original packet which
401GRO has decided not to coalesce, it is placed on a per-NAPI list. This
402list is then passed to the stack when the number of segments reaches the
403gro_normal_batch limit.
404
405high_order_alloc_disable
406------------------------
407
408By default the allocator for page frags tries to use high order pages (order-3
409on x86). While the default behavior gives good results in most cases, some users
410might have hit a contention in page allocations/freeing. This was especially
411true on older kernels (< 5.14) when high-order pages were not stored on per-cpu
412lists. This allows to opt-in for order-0 allocation instead but is now mostly of
413historical importance.
414
415Default: 0
416
4172. /proc/sys/net/unix - Parameters for Unix domain sockets
418----------------------------------------------------------
419
420There is only one file in this directory.
421unix_dgram_qlen limits the max number of datagrams queued in Unix domain
422socket's buffer. It will not take effect unless PF_UNIX flag is specified.
423
424
4253. /proc/sys/net/ipv4 - IPV4 settings
426-------------------------------------
427Please see: Documentation/networking/ip-sysctl.rst and
428Documentation/admin-guide/sysctl/net.rst for descriptions of these entries.
429
430
4314. Appletalk
432------------
433
434The /proc/sys/net/appletalk  directory  holds the Appletalk configuration data
435when Appletalk is loaded. The configurable parameters are:
436
437aarp-expiry-time
438----------------
439
440The amount  of  time  we keep an ARP entry before expiring it. Used to age out
441old hosts.
442
443aarp-resolve-time
444-----------------
445
446The amount of time we will spend trying to resolve an Appletalk address.
447
448aarp-retransmit-limit
449---------------------
450
451The number of times we will retransmit a query before giving up.
452
453aarp-tick-time
454--------------
455
456Controls the rate at which expires are checked.
457
458The directory  /proc/net/appletalk  holds the list of active Appletalk sockets
459on a machine.
460
461The fields  indicate  the DDP type, the local address (in network:node format)
462the remote  address,  the  size of the transmit pending queue, the size of the
463received queue  (bytes waiting for applications to read) the state and the uid
464owning the socket.
465
466/proc/net/atalk_iface lists  all  the  interfaces  configured for appletalk.It
467shows the  name  of the interface, its Appletalk address, the network range on
468that address  (or  network number for phase 1 networks), and the status of the
469interface.
470
471/proc/net/atalk_route lists  each  known  network  route.  It lists the target
472(network) that the route leads to, the router (may be directly connected), the
473route flags, and the device the route is using.
474
4755. TIPC
476-------
477
478tipc_rmem
479---------
480
481The TIPC protocol now has a tunable for the receive memory, similar to the
482tcp_rmem - i.e. a vector of 3 INTEGERs: (min, default, max)
483
484::
485
486    # cat /proc/sys/net/tipc/tipc_rmem
487    4252725 34021800        68043600
488    #
489
490The max value is set to CONN_OVERLOAD_LIMIT, and the default and min values
491are scaled (shifted) versions of that same value.  Note that the min value
492is not at this point in time used in any meaningful way, but the triplet is
493preserved in order to be consistent with things like tcp_rmem.
494
495named_timeout
496-------------
497
498TIPC name table updates are distributed asynchronously in a cluster, without
499any form of transaction handling. This means that different race scenarios are
500possible. One such is that a name withdrawal sent out by one node and received
501by another node may arrive after a second, overlapping name publication already
502has been accepted from a third node, although the conflicting updates
503originally may have been issued in the correct sequential order.
504If named_timeout is nonzero, failed topology updates will be placed on a defer
505queue until another event arrives that clears the error, or until the timeout
506expires. Value is in milliseconds.
507