1.. SPDX-License-Identifier: GPL-2.0
2
3=========
4IP Sysctl
5=========
6
7/proc/sys/net/ipv4/* Variables
8==============================
9
10ip_forward - BOOLEAN
11	- 0 - disabled (default)
12	- not 0 - enabled
13
14	Forward Packets between interfaces.
15
16	This variable is special, its change resets all configuration
17	parameters to their default state (RFC1122 for hosts, RFC1812
18	for routers)
19
20ip_default_ttl - INTEGER
21	Default value of TTL field (Time To Live) for outgoing (but not
22	forwarded) IP packets. Should be between 1 and 255 inclusive.
23	Default: 64 (as recommended by RFC1700)
24
25ip_no_pmtu_disc - INTEGER
26	Disable Path MTU Discovery. If enabled in mode 1 and a
27	fragmentation-required ICMP is received, the PMTU to this
28	destination will be set to the smallest of the old MTU to
29	this destination and min_pmtu (see below). You will need
30	to raise min_pmtu to the smallest interface MTU on your system
31	manually if you want to avoid locally generated fragments.
32
33	In mode 2 incoming Path MTU Discovery messages will be
34	discarded. Outgoing frames are handled the same as in mode 1,
35	implicitly setting IP_PMTUDISC_DONT on every created socket.
36
37	Mode 3 is a hardened pmtu discover mode. The kernel will only
38	accept fragmentation-needed errors if the underlying protocol
39	can verify them besides a plain socket lookup. Current
40	protocols for which pmtu events will be honored are TCP, SCTP
41	and DCCP as they verify e.g. the sequence number or the
42	association. This mode should not be enabled globally but is
43	only intended to secure e.g. name servers in namespaces where
44	TCP path mtu must still work but path MTU information of other
45	protocols should be discarded. If enabled globally this mode
46	could break other protocols.
47
48	Possible values: 0-3
49
50	Default: FALSE
51
52min_pmtu - INTEGER
53	default 552 - minimum Path MTU. Unless this is changed mannually,
54	each cached pmtu will never be lower than this setting.
55
56ip_forward_use_pmtu - BOOLEAN
57	By default we don't trust protocol path MTUs while forwarding
58	because they could be easily forged and can lead to unwanted
59	fragmentation by the router.
60	You only need to enable this if you have user-space software
61	which tries to discover path mtus by itself and depends on the
62	kernel honoring this information. This is normally not the
63	case.
64
65	Default: 0 (disabled)
66
67	Possible values:
68
69	- 0 - disabled
70	- 1 - enabled
71
72fwmark_reflect - BOOLEAN
73	Controls the fwmark of kernel-generated IPv4 reply packets that are not
74	associated with a socket for example, TCP RSTs or ICMP echo replies).
75	If unset, these packets have a fwmark of zero. If set, they have the
76	fwmark of the packet they are replying to.
77
78	Default: 0
79
80fib_multipath_use_neigh - BOOLEAN
81	Use status of existing neighbor entry when determining nexthop for
82	multipath routes. If disabled, neighbor information is not used and
83	packets could be directed to a failed nexthop. Only valid for kernels
84	built with CONFIG_IP_ROUTE_MULTIPATH enabled.
85
86	Default: 0 (disabled)
87
88	Possible values:
89
90	- 0 - disabled
91	- 1 - enabled
92
93fib_multipath_hash_policy - INTEGER
94	Controls which hash policy to use for multipath routes. Only valid
95	for kernels built with CONFIG_IP_ROUTE_MULTIPATH enabled.
96
97	Default: 0 (Layer 3)
98
99	Possible values:
100
101	- 0 - Layer 3
102	- 1 - Layer 4
103	- 2 - Layer 3 or inner Layer 3 if present
104	- 3 - Custom multipath hash. Fields used for multipath hash calculation
105	  are determined by fib_multipath_hash_fields sysctl
106
107fib_multipath_hash_fields - UNSIGNED INTEGER
108	When fib_multipath_hash_policy is set to 3 (custom multipath hash), the
109	fields used for multipath hash calculation are determined by this
110	sysctl.
111
112	This value is a bitmask which enables various fields for multipath hash
113	calculation.
114
115	Possible fields are:
116
117	====== ============================
118	0x0001 Source IP address
119	0x0002 Destination IP address
120	0x0004 IP protocol
121	0x0008 Unused (Flow Label)
122	0x0010 Source port
123	0x0020 Destination port
124	0x0040 Inner source IP address
125	0x0080 Inner destination IP address
126	0x0100 Inner IP protocol
127	0x0200 Inner Flow Label
128	0x0400 Inner source port
129	0x0800 Inner destination port
130	====== ============================
131
132	Default: 0x0007 (source IP, destination IP and IP protocol)
133
134fib_sync_mem - UNSIGNED INTEGER
135	Amount of dirty memory from fib entries that can be backlogged before
136	synchronize_rcu is forced.
137
138	Default: 512kB   Minimum: 64kB   Maximum: 64MB
139
140ip_forward_update_priority - INTEGER
141	Whether to update SKB priority from "TOS" field in IPv4 header after it
142	is forwarded. The new SKB priority is mapped from TOS field value
143	according to an rt_tos2priority table (see e.g. man tc-prio).
144
145	Default: 1 (Update priority.)
146
147	Possible values:
148
149	- 0 - Do not update priority.
150	- 1 - Update priority.
151
152route/max_size - INTEGER
153	Maximum number of routes allowed in the kernel.  Increase
154	this when using large numbers of interfaces and/or routes.
155
156	From linux kernel 3.6 onwards, this is deprecated for ipv4
157	as route cache is no longer used.
158
159neigh/default/gc_thresh1 - INTEGER
160	Minimum number of entries to keep.  Garbage collector will not
161	purge entries if there are fewer than this number.
162
163	Default: 128
164
165neigh/default/gc_thresh2 - INTEGER
166	Threshold when garbage collector becomes more aggressive about
167	purging entries. Entries older than 5 seconds will be cleared
168	when over this number.
169
170	Default: 512
171
172neigh/default/gc_thresh3 - INTEGER
173	Maximum number of non-PERMANENT neighbor entries allowed.  Increase
174	this when using large numbers of interfaces and when communicating
175	with large numbers of directly-connected peers.
176
177	Default: 1024
178
179neigh/default/unres_qlen_bytes - INTEGER
180	The maximum number of bytes which may be used by packets
181	queued for each	unresolved address by other network layers.
182	(added in linux 3.3)
183
184	Setting negative value is meaningless and will return error.
185
186	Default: SK_WMEM_MAX, (same as net.core.wmem_default).
187
188		Exact value depends on architecture and kernel options,
189		but should be enough to allow queuing 256 packets
190		of medium size.
191
192neigh/default/unres_qlen - INTEGER
193	The maximum number of packets which may be queued for each
194	unresolved address by other network layers.
195
196	(deprecated in linux 3.3) : use unres_qlen_bytes instead.
197
198	Prior to linux 3.3, the default value is 3 which may cause
199	unexpected packet loss. The current default value is calculated
200	according to default value of unres_qlen_bytes and true size of
201	packet.
202
203	Default: 101
204
205mtu_expires - INTEGER
206	Time, in seconds, that cached PMTU information is kept.
207
208min_adv_mss - INTEGER
209	The advertised MSS depends on the first hop route MTU, but will
210	never be lower than this setting.
211
212fib_notify_on_flag_change - INTEGER
213        Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/
214        RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed.
215
216        After installing a route to the kernel, user space receives an
217        acknowledgment, which means the route was installed in the kernel,
218        but not necessarily in hardware.
219        It is also possible for a route already installed in hardware to change
220        its action and therefore its flags. For example, a host route that is
221        trapping packets can be "promoted" to perform decapsulation following
222        the installation of an IPinIP/VXLAN tunnel.
223        The notifications will indicate to user-space the state of the route.
224
225        Default: 0 (Do not emit notifications.)
226
227        Possible values:
228
229        - 0 - Do not emit notifications.
230        - 1 - Emit notifications.
231        - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change.
232
233IP Fragmentation:
234
235ipfrag_high_thresh - LONG INTEGER
236	Maximum memory used to reassemble IP fragments.
237
238ipfrag_low_thresh - LONG INTEGER
239	(Obsolete since linux-4.17)
240	Maximum memory used to reassemble IP fragments before the kernel
241	begins to remove incomplete fragment queues to free up resources.
242	The kernel still accepts new fragments for defragmentation.
243
244ipfrag_time - INTEGER
245	Time in seconds to keep an IP fragment in memory.
246
247ipfrag_max_dist - INTEGER
248	ipfrag_max_dist is a non-negative integer value which defines the
249	maximum "disorder" which is allowed among fragments which share a
250	common IP source address. Note that reordering of packets is
251	not unusual, but if a large number of fragments arrive from a source
252	IP address while a particular fragment queue remains incomplete, it
253	probably indicates that one or more fragments belonging to that queue
254	have been lost. When ipfrag_max_dist is positive, an additional check
255	is done on fragments before they are added to a reassembly queue - if
256	ipfrag_max_dist (or more) fragments have arrived from a particular IP
257	address between additions to any IP fragment queue using that source
258	address, it's presumed that one or more fragments in the queue are
259	lost. The existing fragment queue will be dropped, and a new one
260	started. An ipfrag_max_dist value of zero disables this check.
261
262	Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
263	result in unnecessarily dropping fragment queues when normal
264	reordering of packets occurs, which could lead to poor application
265	performance. Using a very large value, e.g. 50000, increases the
266	likelihood of incorrectly reassembling IP fragments that originate
267	from different IP datagrams, which could result in data corruption.
268	Default: 64
269
270bc_forwarding - INTEGER
271	bc_forwarding enables the feature described in rfc1812#section-5.3.5.2
272	and rfc2644. It allows the router to forward directed broadcast.
273	To enable this feature, the 'all' entry and the input interface entry
274	should be set to 1.
275	Default: 0
276
277INET peer storage
278=================
279
280inet_peer_threshold - INTEGER
281	The approximate size of the storage.  Starting from this threshold
282	entries will be thrown aggressively.  This threshold also determines
283	entries' time-to-live and time intervals between garbage collection
284	passes.  More entries, less time-to-live, less GC interval.
285
286inet_peer_minttl - INTEGER
287	Minimum time-to-live of entries.  Should be enough to cover fragment
288	time-to-live on the reassembling side.  This minimum time-to-live  is
289	guaranteed if the pool size is less than inet_peer_threshold.
290	Measured in seconds.
291
292inet_peer_maxttl - INTEGER
293	Maximum time-to-live of entries.  Unused entries will expire after
294	this period of time if there is no memory pressure on the pool (i.e.
295	when the number of entries in the pool is very small).
296	Measured in seconds.
297
298TCP variables
299=============
300
301somaxconn - INTEGER
302	Limit of socket listen() backlog, known in userspace as SOMAXCONN.
303	Defaults to 4096. (Was 128 before linux-5.4)
304	See also tcp_max_syn_backlog for additional tuning for TCP sockets.
305
306tcp_abort_on_overflow - BOOLEAN
307	If listening service is too slow to accept new connections,
308	reset them. Default state is FALSE. It means that if overflow
309	occurred due to a burst, connection will recover. Enable this
310	option _only_ if you are really sure that listening daemon
311	cannot be tuned to accept connections faster. Enabling this
312	option can harm clients of your server.
313
314tcp_adv_win_scale - INTEGER
315	Count buffering overhead as bytes/2^tcp_adv_win_scale
316	(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
317	if it is <= 0.
318
319	Possible values are [-31, 31], inclusive.
320
321	Default: 1
322
323tcp_allowed_congestion_control - STRING
324	Show/set the congestion control choices available to non-privileged
325	processes. The list is a subset of those listed in
326	tcp_available_congestion_control.
327
328	Default is "reno" and the default setting (tcp_congestion_control).
329
330tcp_app_win - INTEGER
331	Reserve max(window/2^tcp_app_win, mss) of window for application
332	buffer. Value 0 is special, it means that nothing is reserved.
333
334	Default: 31
335
336tcp_autocorking - BOOLEAN
337	Enable TCP auto corking :
338	When applications do consecutive small write()/sendmsg() system calls,
339	we try to coalesce these small writes as much as possible, to lower
340	total amount of sent packets. This is done if at least one prior
341	packet for the flow is waiting in Qdisc queues or device transmit
342	queue. Applications can still use TCP_CORK for optimal behavior
343	when they know how/when to uncork their sockets.
344
345	Default : 1
346
347tcp_available_congestion_control - STRING
348	Shows the available congestion control choices that are registered.
349	More congestion control algorithms may be available as modules,
350	but not loaded.
351
352tcp_base_mss - INTEGER
353	The initial value of search_low to be used by the packetization layer
354	Path MTU discovery (MTU probing).  If MTU probing is enabled,
355	this is the initial MSS used by the connection.
356
357tcp_mtu_probe_floor - INTEGER
358	If MTU probing is enabled this caps the minimum MSS used for search_low
359	for the connection.
360
361	Default : 48
362
363tcp_min_snd_mss - INTEGER
364	TCP SYN and SYNACK messages usually advertise an ADVMSS option,
365	as described in RFC 1122 and RFC 6691.
366
367	If this ADVMSS option is smaller than tcp_min_snd_mss,
368	it is silently capped to tcp_min_snd_mss.
369
370	Default : 48 (at least 8 bytes of payload per segment)
371
372tcp_congestion_control - STRING
373	Set the congestion control algorithm to be used for new
374	connections. The algorithm "reno" is always available, but
375	additional choices may be available based on kernel configuration.
376	Default is set as part of kernel configuration.
377	For passive connections, the listener congestion control choice
378	is inherited.
379
380	[see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
381
382tcp_dsack - BOOLEAN
383	Allows TCP to send "duplicate" SACKs.
384
385tcp_early_retrans - INTEGER
386	Tail loss probe (TLP) converts RTOs occurring due to tail
387	losses into fast recovery (draft-ietf-tcpm-rack). Note that
388	TLP requires RACK to function properly (see tcp_recovery below)
389
390	Possible values:
391
392		- 0 disables TLP
393		- 3 or 4 enables TLP
394
395	Default: 3
396
397tcp_ecn - INTEGER
398	Control use of Explicit Congestion Notification (ECN) by TCP.
399	ECN is used only when both ends of the TCP connection indicate
400	support for it.  This feature is useful in avoiding losses due
401	to congestion by allowing supporting routers to signal
402	congestion before having to drop packets.
403
404	Possible values are:
405
406		=  =====================================================
407		0  Disable ECN.  Neither initiate nor accept ECN.
408		1  Enable ECN when requested by incoming connections and
409		   also request ECN on outgoing connection attempts.
410		2  Enable ECN when requested by incoming connections
411		   but do not request ECN on outgoing connections.
412		=  =====================================================
413
414	Default: 2
415
416tcp_ecn_fallback - BOOLEAN
417	If the kernel detects that ECN connection misbehaves, enable fall
418	back to non-ECN. Currently, this knob implements the fallback
419	from RFC3168, section 6.1.1.1., but we reserve that in future,
420	additional detection mechanisms could be implemented under this
421	knob. The value	is not used, if tcp_ecn or per route (or congestion
422	control) ECN settings are disabled.
423
424	Default: 1 (fallback enabled)
425
426tcp_fack - BOOLEAN
427	This is a legacy option, it has no effect anymore.
428
429tcp_fin_timeout - INTEGER
430	The length of time an orphaned (no longer referenced by any
431	application) connection will remain in the FIN_WAIT_2 state
432	before it is aborted at the local end.  While a perfectly
433	valid "receive only" state for an un-orphaned connection, an
434	orphaned connection in FIN_WAIT_2 state could otherwise wait
435	forever for the remote to close its end of the connection.
436
437	Cf. tcp_max_orphans
438
439	Default: 60 seconds
440
441tcp_frto - INTEGER
442	Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
443	F-RTO is an enhanced recovery algorithm for TCP retransmission
444	timeouts.  It is particularly beneficial in networks where the
445	RTT fluctuates (e.g., wireless). F-RTO is sender-side only
446	modification. It does not require any support from the peer.
447
448	By default it's enabled with a non-zero value. 0 disables F-RTO.
449
450tcp_fwmark_accept - BOOLEAN
451	If set, incoming connections to listening sockets that do not have a
452	socket mark will set the mark of the accepting socket to the fwmark of
453	the incoming SYN packet. This will cause all packets on that connection
454	(starting from the first SYNACK) to be sent with that fwmark. The
455	listening socket's mark is unchanged. Listening sockets that already
456	have a fwmark set via setsockopt(SOL_SOCKET, SO_MARK, ...) are
457	unaffected.
458
459	Default: 0
460
461tcp_invalid_ratelimit - INTEGER
462	Limit the maximal rate for sending duplicate acknowledgments
463	in response to incoming TCP packets that are for an existing
464	connection but that are invalid due to any of these reasons:
465
466	  (a) out-of-window sequence number,
467	  (b) out-of-window acknowledgment number, or
468	  (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
469
470	This can help mitigate simple "ack loop" DoS attacks, wherein
471	a buggy or malicious middlebox or man-in-the-middle can
472	rewrite TCP header fields in manner that causes each endpoint
473	to think that the other is sending invalid TCP segments, thus
474	causing each side to send an unterminating stream of duplicate
475	acknowledgments for invalid segments.
476
477	Using 0 disables rate-limiting of dupacks in response to
478	invalid segments; otherwise this value specifies the minimal
479	space between sending such dupacks, in milliseconds.
480
481	Default: 500 (milliseconds).
482
483tcp_keepalive_time - INTEGER
484	How often TCP sends out keepalive messages when keepalive is enabled.
485	Default: 2hours.
486
487tcp_keepalive_probes - INTEGER
488	How many keepalive probes TCP sends out, until it decides that the
489	connection is broken. Default value: 9.
490
491tcp_keepalive_intvl - INTEGER
492	How frequently the probes are send out. Multiplied by
493	tcp_keepalive_probes it is time to kill not responding connection,
494	after probes started. Default value: 75sec i.e. connection
495	will be aborted after ~11 minutes of retries.
496
497tcp_l3mdev_accept - BOOLEAN
498	Enables child sockets to inherit the L3 master device index.
499	Enabling this option allows a "global" listen socket to work
500	across L3 master domains (e.g., VRFs) with connected sockets
501	derived from the listen socket to be bound to the L3 domain in
502	which the packets originated. Only valid when the kernel was
503	compiled with CONFIG_NET_L3_MASTER_DEV.
504
505	Default: 0 (disabled)
506
507tcp_low_latency - BOOLEAN
508	This is a legacy option, it has no effect anymore.
509
510tcp_max_orphans - INTEGER
511	Maximal number of TCP sockets not attached to any user file handle,
512	held by system.	If this number is exceeded orphaned connections are
513	reset immediately and warning is printed. This limit exists
514	only to prevent simple DoS attacks, you _must_ not rely on this
515	or lower the limit artificially, but rather increase it
516	(probably, after increasing installed memory),
517	if network conditions require more than default value,
518	and tune network services to linger and kill such states
519	more aggressively. Let me to remind again: each orphan eats
520	up to ~64K of unswappable memory.
521
522tcp_max_syn_backlog - INTEGER
523	Maximal number of remembered connection requests (SYN_RECV),
524	which have not received an acknowledgment from connecting client.
525
526	This is a per-listener limit.
527
528	The minimal value is 128 for low memory machines, and it will
529	increase in proportion to the memory of machine.
530
531	If server suffers from overload, try increasing this number.
532
533	Remember to also check /proc/sys/net/core/somaxconn
534	A SYN_RECV request socket consumes about 304 bytes of memory.
535
536tcp_max_tw_buckets - INTEGER
537	Maximal number of timewait sockets held by system simultaneously.
538	If this number is exceeded time-wait socket is immediately destroyed
539	and warning is printed. This limit exists only to prevent
540	simple DoS attacks, you _must_ not lower the limit artificially,
541	but rather increase it (probably, after increasing installed memory),
542	if network conditions require more than default value.
543
544tcp_mem - vector of 3 INTEGERs: min, pressure, max
545	min: below this number of pages TCP is not bothered about its
546	memory appetite.
547
548	pressure: when amount of memory allocated by TCP exceeds this number
549	of pages, TCP moderates its memory consumption and enters memory
550	pressure mode, which is exited when memory consumption falls
551	under "min".
552
553	max: number of pages allowed for queueing by all TCP sockets.
554
555	Defaults are calculated at boot time from amount of available
556	memory.
557
558tcp_min_rtt_wlen - INTEGER
559	The window length of the windowed min filter to track the minimum RTT.
560	A shorter window lets a flow more quickly pick up new (higher)
561	minimum RTT when it is moved to a longer path (e.g., due to traffic
562	engineering). A longer window makes the filter more resistant to RTT
563	inflations such as transient congestion. The unit is seconds.
564
565	Possible values: 0 - 86400 (1 day)
566
567	Default: 300
568
569tcp_moderate_rcvbuf - BOOLEAN
570	If set, TCP performs receive buffer auto-tuning, attempting to
571	automatically size the buffer (no greater than tcp_rmem[2]) to
572	match the size required by the path for full throughput.  Enabled by
573	default.
574
575tcp_mtu_probing - INTEGER
576	Controls TCP Packetization-Layer Path MTU Discovery.  Takes three
577	values:
578
579	- 0 - Disabled
580	- 1 - Disabled by default, enabled when an ICMP black hole detected
581	- 2 - Always enabled, use initial MSS of tcp_base_mss.
582
583tcp_probe_interval - UNSIGNED INTEGER
584	Controls how often to start TCP Packetization-Layer Path MTU
585	Discovery reprobe. The default is reprobing every 10 minutes as
586	per RFC4821.
587
588tcp_probe_threshold - INTEGER
589	Controls when TCP Packetization-Layer Path MTU Discovery probing
590	will stop in respect to the width of search range in bytes. Default
591	is 8 bytes.
592
593tcp_no_metrics_save - BOOLEAN
594	By default, TCP saves various connection metrics in the route cache
595	when the connection closes, so that connections established in the
596	near future can use these to set initial conditions.  Usually, this
597	increases overall performance, but may sometimes cause performance
598	degradation.  If set, TCP will not cache metrics on closing
599	connections.
600
601tcp_no_ssthresh_metrics_save - BOOLEAN
602	Controls whether TCP saves ssthresh metrics in the route cache.
603
604	Default is 1, which disables ssthresh metrics.
605
606tcp_orphan_retries - INTEGER
607	This value influences the timeout of a locally closed TCP connection,
608	when RTO retransmissions remain unacknowledged.
609	See tcp_retries2 for more details.
610
611	The default value is 8.
612
613	If your machine is a loaded WEB server,
614	you should think about lowering this value, such sockets
615	may consume significant resources. Cf. tcp_max_orphans.
616
617tcp_recovery - INTEGER
618	This value is a bitmap to enable various experimental loss recovery
619	features.
620
621	=========   =============================================================
622	RACK: 0x1   enables the RACK loss detection for fast detection of lost
623		    retransmissions and tail drops. It also subsumes and disables
624		    RFC6675 recovery for SACK connections.
625
626	RACK: 0x2   makes RACK's reordering window static (min_rtt/4).
627
628	RACK: 0x4   disables RACK's DUPACK threshold heuristic
629	=========   =============================================================
630
631	Default: 0x1
632
633tcp_reordering - INTEGER
634	Initial reordering level of packets in a TCP stream.
635	TCP stack can then dynamically adjust flow reordering level
636	between this initial value and tcp_max_reordering
637
638	Default: 3
639
640tcp_max_reordering - INTEGER
641	Maximal reordering level of packets in a TCP stream.
642	300 is a fairly conservative value, but you might increase it
643	if paths are using per packet load balancing (like bonding rr mode)
644
645	Default: 300
646
647tcp_retrans_collapse - BOOLEAN
648	Bug-to-bug compatibility with some broken printers.
649	On retransmit try to send bigger packets to work around bugs in
650	certain TCP stacks.
651
652tcp_retries1 - INTEGER
653	This value influences the time, after which TCP decides, that
654	something is wrong due to unacknowledged RTO retransmissions,
655	and reports this suspicion to the network layer.
656	See tcp_retries2 for more details.
657
658	RFC 1122 recommends at least 3 retransmissions, which is the
659	default.
660
661tcp_retries2 - INTEGER
662	This value influences the timeout of an alive TCP connection,
663	when RTO retransmissions remain unacknowledged.
664	Given a value of N, a hypothetical TCP connection following
665	exponential backoff with an initial RTO of TCP_RTO_MIN would
666	retransmit N times before killing the connection at the (N+1)th RTO.
667
668	The default value of 15 yields a hypothetical timeout of 924.6
669	seconds and is a lower bound for the effective timeout.
670	TCP will effectively time out at the first RTO which exceeds the
671	hypothetical timeout.
672
673	RFC 1122 recommends at least 100 seconds for the timeout,
674	which corresponds to a value of at least 8.
675
676tcp_rfc1337 - BOOLEAN
677	If set, the TCP stack behaves conforming to RFC1337. If unset,
678	we are not conforming to RFC, but prevent TCP TIME_WAIT
679	assassination.
680
681	Default: 0
682
683tcp_rmem - vector of 3 INTEGERs: min, default, max
684	min: Minimal size of receive buffer used by TCP sockets.
685	It is guaranteed to each TCP socket, even under moderate memory
686	pressure.
687
688	Default: 4K
689
690	default: initial size of receive buffer used by TCP sockets.
691	This value overrides net.core.rmem_default used by other protocols.
692	Default: 131072 bytes.
693	This value results in initial window of 65535.
694
695	max: maximal size of receive buffer allowed for automatically
696	selected receiver buffers for TCP socket. This value does not override
697	net.core.rmem_max.  Calling setsockopt() with SO_RCVBUF disables
698	automatic tuning of that socket's receive buffer size, in which
699	case this value is ignored.
700	Default: between 131072 and 6MB, depending on RAM size.
701
702tcp_sack - BOOLEAN
703	Enable select acknowledgments (SACKS).
704
705tcp_comp_sack_delay_ns - LONG INTEGER
706	TCP tries to reduce number of SACK sent, using a timer
707	based on 5% of SRTT, capped by this sysctl, in nano seconds.
708	The default is 1ms, based on TSO autosizing period.
709
710	Default : 1,000,000 ns (1 ms)
711
712tcp_comp_sack_slack_ns - LONG INTEGER
713	This sysctl control the slack used when arming the
714	timer used by SACK compression. This gives extra time
715	for small RTT flows, and reduces system overhead by allowing
716	opportunistic reduction of timer interrupts.
717
718	Default : 100,000 ns (100 us)
719
720tcp_comp_sack_nr - INTEGER
721	Max number of SACK that can be compressed.
722	Using 0 disables SACK compression.
723
724	Default : 44
725
726tcp_slow_start_after_idle - BOOLEAN
727	If set, provide RFC2861 behavior and time out the congestion
728	window after an idle period.  An idle period is defined at
729	the current RTO.  If unset, the congestion window will not
730	be timed out after an idle period.
731
732	Default: 1
733
734tcp_stdurg - BOOLEAN
735	Use the Host requirements interpretation of the TCP urgent pointer field.
736	Most hosts use the older BSD interpretation, so if you turn this on
737	Linux might not communicate correctly with them.
738
739	Default: FALSE
740
741tcp_synack_retries - INTEGER
742	Number of times SYNACKs for a passive TCP connection attempt will
743	be retransmitted. Should not be higher than 255. Default value
744	is 5, which corresponds to 31seconds till the last retransmission
745	with the current initial RTO of 1second. With this the final timeout
746	for a passive TCP connection will happen after 63seconds.
747
748tcp_syncookies - INTEGER
749	Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
750	Send out syncookies when the syn backlog queue of a socket
751	overflows. This is to prevent against the common 'SYN flood attack'
752	Default: 1
753
754	Note, that syncookies is fallback facility.
755	It MUST NOT be used to help highly loaded servers to stand
756	against legal connection rate. If you see SYN flood warnings
757	in your logs, but investigation	shows that they occur
758	because of overload with legal connections, you should tune
759	another parameters until this warning disappear.
760	See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
761
762	syncookies seriously violate TCP protocol, do not allow
763	to use TCP extensions, can result in serious degradation
764	of some services (f.e. SMTP relaying), visible not by you,
765	but your clients and relays, contacting you. While you see
766	SYN flood warnings in logs not being really flooded, your server
767	is seriously misconfigured.
768
769	If you want to test which effects syncookies have to your
770	network connections you can set this knob to 2 to enable
771	unconditionally generation of syncookies.
772
773tcp_migrate_req - BOOLEAN
774	The incoming connection is tied to a specific listening socket when
775	the initial SYN packet is received during the three-way handshake.
776	When a listener is closed, in-flight request sockets during the
777	handshake and established sockets in the accept queue are aborted.
778
779	If the listener has SO_REUSEPORT enabled, other listeners on the
780	same port should have been able to accept such connections. This
781	option makes it possible to migrate such child sockets to another
782	listener after close() or shutdown().
783
784	The BPF_SK_REUSEPORT_SELECT_OR_MIGRATE type of eBPF program should
785	usually be used to define the policy to pick an alive listener.
786	Otherwise, the kernel will randomly pick an alive listener only if
787	this option is enabled.
788
789	Note that migration between listeners with different settings may
790	crash applications. Let's say migration happens from listener A to
791	B, and only B has TCP_SAVE_SYN enabled. B cannot read SYN data from
792	the requests migrated from A. To avoid such a situation, cancel
793	migration by returning SK_DROP in the type of eBPF program, or
794	disable this option.
795
796	Default: 0
797
798tcp_fastopen - INTEGER
799	Enable TCP Fast Open (RFC7413) to send and accept data in the opening
800	SYN packet.
801
802	The client support is enabled by flag 0x1 (on by default). The client
803	then must use sendmsg() or sendto() with the MSG_FASTOPEN flag,
804	rather than connect() to send data in SYN.
805
806	The server support is enabled by flag 0x2 (off by default). Then
807	either enable for all listeners with another flag (0x400) or
808	enable individual listeners via TCP_FASTOPEN socket option with
809	the option value being the length of the syn-data backlog.
810
811	The values (bitmap) are
812
813	=====  ======== ======================================================
814	  0x1  (client) enables sending data in the opening SYN on the client.
815	  0x2  (server) enables the server support, i.e., allowing data in
816			a SYN packet to be accepted and passed to the
817			application before 3-way handshake finishes.
818	  0x4  (client) send data in the opening SYN regardless of cookie
819			availability and without a cookie option.
820	0x200  (server) accept data-in-SYN w/o any cookie option present.
821	0x400  (server) enable all listeners to support Fast Open by
822			default without explicit TCP_FASTOPEN socket option.
823	=====  ======== ======================================================
824
825	Default: 0x1
826
827	Note that additional client or server features are only
828	effective if the basic support (0x1 and 0x2) are enabled respectively.
829
830tcp_fastopen_blackhole_timeout_sec - INTEGER
831	Initial time period in second to disable Fastopen on active TCP sockets
832	when a TFO firewall blackhole issue happens.
833	This time period will grow exponentially when more blackhole issues
834	get detected right after Fastopen is re-enabled and will reset to
835	initial value when the blackhole issue goes away.
836	0 to disable the blackhole detection.
837
838	By default, it is set to 0 (feature is disabled).
839
840tcp_fastopen_key - list of comma separated 32-digit hexadecimal INTEGERs
841	The list consists of a primary key and an optional backup key. The
842	primary key is used for both creating and validating cookies, while the
843	optional backup key is only used for validating cookies. The purpose of
844	the backup key is to maximize TFO validation when keys are rotated.
845
846	A randomly chosen primary key may be configured by the kernel if
847	the tcp_fastopen sysctl is set to 0x400 (see above), or if the
848	TCP_FASTOPEN setsockopt() optname is set and a key has not been
849	previously configured via sysctl. If keys are configured via
850	setsockopt() by using the TCP_FASTOPEN_KEY optname, then those
851	per-socket keys will be used instead of any keys that are specified via
852	sysctl.
853
854	A key is specified as 4 8-digit hexadecimal integers which are separated
855	by a '-' as: xxxxxxxx-xxxxxxxx-xxxxxxxx-xxxxxxxx. Leading zeros may be
856	omitted. A primary and a backup key may be specified by separating them
857	by a comma. If only one key is specified, it becomes the primary key and
858	any previously configured backup keys are removed.
859
860tcp_syn_retries - INTEGER
861	Number of times initial SYNs for an active TCP connection attempt
862	will be retransmitted. Should not be higher than 127. Default value
863	is 6, which corresponds to 63seconds till the last retransmission
864	with the current initial RTO of 1second. With this the final timeout
865	for an active TCP connection attempt will happen after 127seconds.
866
867tcp_timestamps - INTEGER
868	Enable timestamps as defined in RFC1323.
869
870	- 0: Disabled.
871	- 1: Enable timestamps as defined in RFC1323 and use random offset for
872	  each connection rather than only using the current time.
873	- 2: Like 1, but without random offsets.
874
875	Default: 1
876
877tcp_min_tso_segs - INTEGER
878	Minimal number of segments per TSO frame.
879
880	Since linux-3.12, TCP does an automatic sizing of TSO frames,
881	depending on flow rate, instead of filling 64Kbytes packets.
882	For specific usages, it's possible to force TCP to build big
883	TSO frames. Note that TCP stack might split too big TSO packets
884	if available window is too small.
885
886	Default: 2
887
888tcp_tso_rtt_log - INTEGER
889	Adjustment of TSO packet sizes based on min_rtt
890
891	Starting from linux-5.18, TCP autosizing can be tweaked
892	for flows having small RTT.
893
894	Old autosizing was splitting the pacing budget to send 1024 TSO
895	per second.
896
897	tso_packet_size = sk->sk_pacing_rate / 1024;
898
899	With the new mechanism, we increase this TSO sizing using:
900
901	distance = min_rtt_usec / (2^tcp_tso_rtt_log)
902	tso_packet_size += gso_max_size >> distance;
903
904	This means that flows between very close hosts can use bigger
905	TSO packets, reducing their cpu costs.
906
907	If you want to use the old autosizing, set this sysctl to 0.
908
909	Default: 9  (2^9 = 512 usec)
910
911tcp_pacing_ss_ratio - INTEGER
912	sk->sk_pacing_rate is set by TCP stack using a ratio applied
913	to current rate. (current_rate = cwnd * mss / srtt)
914	If TCP is in slow start, tcp_pacing_ss_ratio is applied
915	to let TCP probe for bigger speeds, assuming cwnd can be
916	doubled every other RTT.
917
918	Default: 200
919
920tcp_pacing_ca_ratio - INTEGER
921	sk->sk_pacing_rate is set by TCP stack using a ratio applied
922	to current rate. (current_rate = cwnd * mss / srtt)
923	If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio
924	is applied to conservatively probe for bigger throughput.
925
926	Default: 120
927
928tcp_tso_win_divisor - INTEGER
929	This allows control over what percentage of the congestion window
930	can be consumed by a single TSO frame.
931	The setting of this parameter is a choice between burstiness and
932	building larger TSO frames.
933
934	Default: 3
935
936tcp_tw_reuse - INTEGER
937	Enable reuse of TIME-WAIT sockets for new connections when it is
938	safe from protocol viewpoint.
939
940	- 0 - disable
941	- 1 - global enable
942	- 2 - enable for loopback traffic only
943
944	It should not be changed without advice/request of technical
945	experts.
946
947	Default: 2
948
949tcp_window_scaling - BOOLEAN
950	Enable window scaling as defined in RFC1323.
951
952tcp_wmem - vector of 3 INTEGERs: min, default, max
953	min: Amount of memory reserved for send buffers for TCP sockets.
954	Each TCP socket has rights to use it due to fact of its birth.
955
956	Default: 4K
957
958	default: initial size of send buffer used by TCP sockets.  This
959	value overrides net.core.wmem_default used by other protocols.
960
961	It is usually lower than net.core.wmem_default.
962
963	Default: 16K
964
965	max: Maximal amount of memory allowed for automatically tuned
966	send buffers for TCP sockets. This value does not override
967	net.core.wmem_max.  Calling setsockopt() with SO_SNDBUF disables
968	automatic tuning of that socket's send buffer size, in which case
969	this value is ignored.
970
971	Default: between 64K and 4MB, depending on RAM size.
972
973tcp_notsent_lowat - UNSIGNED INTEGER
974	A TCP socket can control the amount of unsent bytes in its write queue,
975	thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
976	reports POLLOUT events if the amount of unsent bytes is below a per
977	socket value, and if the write queue is not full. sendmsg() will
978	also not add new buffers if the limit is hit.
979
980	This global variable controls the amount of unsent data for
981	sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
982	to the global variable has immediate effect.
983
984	Default: UINT_MAX (0xFFFFFFFF)
985
986tcp_workaround_signed_windows - BOOLEAN
987	If set, assume no receipt of a window scaling option means the
988	remote TCP is broken and treats the window as a signed quantity.
989	If unset, assume the remote TCP is not broken even if we do
990	not receive a window scaling option from them.
991
992	Default: 0
993
994tcp_thin_linear_timeouts - BOOLEAN
995	Enable dynamic triggering of linear timeouts for thin streams.
996	If set, a check is performed upon retransmission by timeout to
997	determine if the stream is thin (less than 4 packets in flight).
998	As long as the stream is found to be thin, up to 6 linear
999	timeouts may be performed before exponential backoff mode is
1000	initiated. This improves retransmission latency for
1001	non-aggressive thin streams, often found to be time-dependent.
1002	For more information on thin streams, see
1003	Documentation/networking/tcp-thin.rst
1004
1005	Default: 0
1006
1007tcp_limit_output_bytes - INTEGER
1008	Controls TCP Small Queue limit per tcp socket.
1009	TCP bulk sender tends to increase packets in flight until it
1010	gets losses notifications. With SNDBUF autotuning, this can
1011	result in a large amount of packets queued on the local machine
1012	(e.g.: qdiscs, CPU backlog, or device) hurting latency of other
1013	flows, for typical pfifo_fast qdiscs.  tcp_limit_output_bytes
1014	limits the number of bytes on qdisc or device to reduce artificial
1015	RTT/cwnd and reduce bufferbloat.
1016
1017	Default: 1048576 (16 * 65536)
1018
1019tcp_challenge_ack_limit - INTEGER
1020	Limits number of Challenge ACK sent per second, as recommended
1021	in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
1022	Default: 1000
1023
1024UDP variables
1025=============
1026
1027udp_l3mdev_accept - BOOLEAN
1028	Enabling this option allows a "global" bound socket to work
1029	across L3 master domains (e.g., VRFs) with packets capable of
1030	being received regardless of the L3 domain in which they
1031	originated. Only valid when the kernel was compiled with
1032	CONFIG_NET_L3_MASTER_DEV.
1033
1034	Default: 0 (disabled)
1035
1036udp_mem - vector of 3 INTEGERs: min, pressure, max
1037	Number of pages allowed for queueing by all UDP sockets.
1038
1039	min: Number of pages allowed for queueing by all UDP sockets.
1040
1041	pressure: This value was introduced to follow format of tcp_mem.
1042
1043	max: This value was introduced to follow format of tcp_mem.
1044
1045	Default is calculated at boot time from amount of available memory.
1046
1047udp_rmem_min - INTEGER
1048	Minimal size of receive buffer used by UDP sockets in moderation.
1049	Each UDP socket is able to use the size for receiving data, even if
1050	total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
1051
1052	Default: 4K
1053
1054udp_wmem_min - INTEGER
1055	Minimal size of send buffer used by UDP sockets in moderation.
1056	Each UDP socket is able to use the size for sending data, even if
1057	total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
1058
1059	Default: 4K
1060
1061RAW variables
1062=============
1063
1064raw_l3mdev_accept - BOOLEAN
1065	Enabling this option allows a "global" bound socket to work
1066	across L3 master domains (e.g., VRFs) with packets capable of
1067	being received regardless of the L3 domain in which they
1068	originated. Only valid when the kernel was compiled with
1069	CONFIG_NET_L3_MASTER_DEV.
1070
1071	Default: 1 (enabled)
1072
1073CIPSOv4 Variables
1074=================
1075
1076cipso_cache_enable - BOOLEAN
1077	If set, enable additions to and lookups from the CIPSO label mapping
1078	cache.  If unset, additions are ignored and lookups always result in a
1079	miss.  However, regardless of the setting the cache is still
1080	invalidated when required when means you can safely toggle this on and
1081	off and the cache will always be "safe".
1082
1083	Default: 1
1084
1085cipso_cache_bucket_size - INTEGER
1086	The CIPSO label cache consists of a fixed size hash table with each
1087	hash bucket containing a number of cache entries.  This variable limits
1088	the number of entries in each hash bucket; the larger the value the
1089	more CIPSO label mappings that can be cached.  When the number of
1090	entries in a given hash bucket reaches this limit adding new entries
1091	causes the oldest entry in the bucket to be removed to make room.
1092
1093	Default: 10
1094
1095cipso_rbm_optfmt - BOOLEAN
1096	Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
1097	the CIPSO draft specification (see Documentation/netlabel for details).
1098	This means that when set the CIPSO tag will be padded with empty
1099	categories in order to make the packet data 32-bit aligned.
1100
1101	Default: 0
1102
1103cipso_rbm_structvalid - BOOLEAN
1104	If set, do a very strict check of the CIPSO option when
1105	ip_options_compile() is called.  If unset, relax the checks done during
1106	ip_options_compile().  Either way is "safe" as errors are caught else
1107	where in the CIPSO processing code but setting this to 0 (False) should
1108	result in less work (i.e. it should be faster) but could cause problems
1109	with other implementations that require strict checking.
1110
1111	Default: 0
1112
1113IP Variables
1114============
1115
1116ip_local_port_range - 2 INTEGERS
1117	Defines the local port range that is used by TCP and UDP to
1118	choose the local port. The first number is the first, the
1119	second the last local port number.
1120	If possible, it is better these numbers have different parity
1121	(one even and one odd value).
1122	Must be greater than or equal to ip_unprivileged_port_start.
1123	The default values are 32768 and 60999 respectively.
1124
1125ip_local_reserved_ports - list of comma separated ranges
1126	Specify the ports which are reserved for known third-party
1127	applications. These ports will not be used by automatic port
1128	assignments (e.g. when calling connect() or bind() with port
1129	number 0). Explicit port allocation behavior is unchanged.
1130
1131	The format used for both input and output is a comma separated
1132	list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
1133	10). Writing to the file will clear all previously reserved
1134	ports and update the current list with the one given in the
1135	input.
1136
1137	Note that ip_local_port_range and ip_local_reserved_ports
1138	settings are independent and both are considered by the kernel
1139	when determining which ports are available for automatic port
1140	assignments.
1141
1142	You can reserve ports which are not in the current
1143	ip_local_port_range, e.g.::
1144
1145	    $ cat /proc/sys/net/ipv4/ip_local_port_range
1146	    32000	60999
1147	    $ cat /proc/sys/net/ipv4/ip_local_reserved_ports
1148	    8080,9148
1149
1150	although this is redundant. However such a setting is useful
1151	if later the port range is changed to a value that will
1152	include the reserved ports. Also keep in mind, that overlapping
1153	of these ranges may affect probability of selecting ephemeral
1154	ports which are right after block of reserved ports.
1155
1156	Default: Empty
1157
1158ip_unprivileged_port_start - INTEGER
1159	This is a per-namespace sysctl.  It defines the first
1160	unprivileged port in the network namespace.  Privileged ports
1161	require root or CAP_NET_BIND_SERVICE in order to bind to them.
1162	To disable all privileged ports, set this to 0.  They must not
1163	overlap with the ip_local_port_range.
1164
1165	Default: 1024
1166
1167ip_nonlocal_bind - BOOLEAN
1168	If set, allows processes to bind() to non-local IP addresses,
1169	which can be quite useful - but may break some applications.
1170
1171	Default: 0
1172
1173ip_autobind_reuse - BOOLEAN
1174	By default, bind() does not select the ports automatically even if
1175	the new socket and all sockets bound to the port have SO_REUSEADDR.
1176	ip_autobind_reuse allows bind() to reuse the port and this is useful
1177	when you use bind()+connect(), but may break some applications.
1178	The preferred solution is to use IP_BIND_ADDRESS_NO_PORT and this
1179	option should only be set by experts.
1180	Default: 0
1181
1182ip_dynaddr - BOOLEAN
1183	If set non-zero, enables support for dynamic addresses.
1184	If set to a non-zero value larger than 1, a kernel log
1185	message will be printed when dynamic address rewriting
1186	occurs.
1187
1188	Default: 0
1189
1190ip_early_demux - BOOLEAN
1191	Optimize input packet processing down to one demux for
1192	certain kinds of local sockets.  Currently we only do this
1193	for established TCP and connected UDP sockets.
1194
1195	It may add an additional cost for pure routing workloads that
1196	reduces overall throughput, in such case you should disable it.
1197
1198	Default: 1
1199
1200ping_group_range - 2 INTEGERS
1201	Restrict ICMP_PROTO datagram sockets to users in the group range.
1202	The default is "1 0", meaning, that nobody (not even root) may
1203	create ping sockets.  Setting it to "100 100" would grant permissions
1204	to the single group. "0 4294967295" would enable it for the world, "100
1205	4294967295" would enable it for the users, but not daemons.
1206
1207tcp_early_demux - BOOLEAN
1208	Enable early demux for established TCP sockets.
1209
1210	Default: 1
1211
1212udp_early_demux - BOOLEAN
1213	Enable early demux for connected UDP sockets. Disable this if
1214	your system could experience more unconnected load.
1215
1216	Default: 1
1217
1218icmp_echo_ignore_all - BOOLEAN
1219	If set non-zero, then the kernel will ignore all ICMP ECHO
1220	requests sent to it.
1221
1222	Default: 0
1223
1224icmp_echo_enable_probe - BOOLEAN
1225        If set to one, then the kernel will respond to RFC 8335 PROBE
1226        requests sent to it.
1227
1228        Default: 0
1229
1230icmp_echo_ignore_broadcasts - BOOLEAN
1231	If set non-zero, then the kernel will ignore all ICMP ECHO and
1232	TIMESTAMP requests sent to it via broadcast/multicast.
1233
1234	Default: 1
1235
1236icmp_ratelimit - INTEGER
1237	Limit the maximal rates for sending ICMP packets whose type matches
1238	icmp_ratemask (see below) to specific targets.
1239	0 to disable any limiting,
1240	otherwise the minimal space between responses in milliseconds.
1241	Note that another sysctl, icmp_msgs_per_sec limits the number
1242	of ICMP packets	sent on all targets.
1243
1244	Default: 1000
1245
1246icmp_msgs_per_sec - INTEGER
1247	Limit maximal number of ICMP packets sent per second from this host.
1248	Only messages whose type matches icmp_ratemask (see below) are
1249	controlled by this limit. For security reasons, the precise count
1250	of messages per second is randomized.
1251
1252	Default: 1000
1253
1254icmp_msgs_burst - INTEGER
1255	icmp_msgs_per_sec controls number of ICMP packets sent per second,
1256	while icmp_msgs_burst controls the burst size of these packets.
1257	For security reasons, the precise burst size is randomized.
1258
1259	Default: 50
1260
1261icmp_ratemask - INTEGER
1262	Mask made of ICMP types for which rates are being limited.
1263
1264	Significant bits: IHGFEDCBA9876543210
1265
1266	Default mask:     0000001100000011000 (6168)
1267
1268	Bit definitions (see include/linux/icmp.h):
1269
1270		= =========================
1271		0 Echo Reply
1272		3 Destination Unreachable [1]_
1273		4 Source Quench [1]_
1274		5 Redirect
1275		8 Echo Request
1276		B Time Exceeded [1]_
1277		C Parameter Problem [1]_
1278		D Timestamp Request
1279		E Timestamp Reply
1280		F Info Request
1281		G Info Reply
1282		H Address Mask Request
1283		I Address Mask Reply
1284		= =========================
1285
1286	.. [1] These are rate limited by default (see default mask above)
1287
1288icmp_ignore_bogus_error_responses - BOOLEAN
1289	Some routers violate RFC1122 by sending bogus responses to broadcast
1290	frames.  Such violations are normally logged via a kernel warning.
1291	If this is set to TRUE, the kernel will not give such warnings, which
1292	will avoid log file clutter.
1293
1294	Default: 1
1295
1296icmp_errors_use_inbound_ifaddr - BOOLEAN
1297
1298	If zero, icmp error messages are sent with the primary address of
1299	the exiting interface.
1300
1301	If non-zero, the message will be sent with the primary address of
1302	the interface that received the packet that caused the icmp error.
1303	This is the behaviour many network administrators will expect from
1304	a router. And it can make debugging complicated network layouts
1305	much easier.
1306
1307	Note that if no primary address exists for the interface selected,
1308	then the primary address of the first non-loopback interface that
1309	has one will be used regardless of this setting.
1310
1311	Default: 0
1312
1313igmp_max_memberships - INTEGER
1314	Change the maximum number of multicast groups we can subscribe to.
1315	Default: 20
1316
1317	Theoretical maximum value is bounded by having to send a membership
1318	report in a single datagram (i.e. the report can't span multiple
1319	datagrams, or risk confusing the switch and leaving groups you don't
1320	intend to).
1321
1322	The number of supported groups 'M' is bounded by the number of group
1323	report entries you can fit into a single datagram of 65535 bytes.
1324
1325	M = 65536-sizeof (ip header)/(sizeof(Group record))
1326
1327	Group records are variable length, with a minimum of 12 bytes.
1328	So net.ipv4.igmp_max_memberships should not be set higher than:
1329
1330	(65536-24) / 12 = 5459
1331
1332	The value 5459 assumes no IP header options, so in practice
1333	this number may be lower.
1334
1335igmp_max_msf - INTEGER
1336	Maximum number of addresses allowed in the source filter list for a
1337	multicast group.
1338
1339	Default: 10
1340
1341igmp_qrv - INTEGER
1342	Controls the IGMP query robustness variable (see RFC2236 8.1).
1343
1344	Default: 2 (as specified by RFC2236 8.1)
1345
1346	Minimum: 1 (as specified by RFC6636 4.5)
1347
1348force_igmp_version - INTEGER
1349	- 0 - (default) No enforcement of a IGMP version, IGMPv1/v2 fallback
1350	  allowed. Will back to IGMPv3 mode again if all IGMPv1/v2 Querier
1351	  Present timer expires.
1352	- 1 - Enforce to use IGMP version 1. Will also reply IGMPv1 report if
1353	  receive IGMPv2/v3 query.
1354	- 2 - Enforce to use IGMP version 2. Will fallback to IGMPv1 if receive
1355	  IGMPv1 query message. Will reply report if receive IGMPv3 query.
1356	- 3 - Enforce to use IGMP version 3. The same react with default 0.
1357
1358	.. note::
1359
1360	   this is not the same with force_mld_version because IGMPv3 RFC3376
1361	   Security Considerations does not have clear description that we could
1362	   ignore other version messages completely as MLDv2 RFC3810. So make
1363	   this value as default 0 is recommended.
1364
1365``conf/interface/*``
1366	changes special settings per interface (where
1367	interface" is the name of your network interface)
1368
1369``conf/all/*``
1370	  is special, changes the settings for all interfaces
1371
1372log_martians - BOOLEAN
1373	Log packets with impossible addresses to kernel log.
1374	log_martians for the interface will be enabled if at least one of
1375	conf/{all,interface}/log_martians is set to TRUE,
1376	it will be disabled otherwise
1377
1378accept_redirects - BOOLEAN
1379	Accept ICMP redirect messages.
1380	accept_redirects for the interface will be enabled if:
1381
1382	- both conf/{all,interface}/accept_redirects are TRUE in the case
1383	  forwarding for the interface is enabled
1384
1385	or
1386
1387	- at least one of conf/{all,interface}/accept_redirects is TRUE in the
1388	  case forwarding for the interface is disabled
1389
1390	accept_redirects for the interface will be disabled otherwise
1391
1392	default:
1393
1394		- TRUE (host)
1395		- FALSE (router)
1396
1397forwarding - BOOLEAN
1398	Enable IP forwarding on this interface.  This controls whether packets
1399	received _on_ this interface can be forwarded.
1400
1401mc_forwarding - BOOLEAN
1402	Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
1403	and a multicast routing daemon is required.
1404	conf/all/mc_forwarding must also be set to TRUE to enable multicast
1405	routing	for the interface
1406
1407medium_id - INTEGER
1408	Integer value used to differentiate the devices by the medium they
1409	are attached to. Two devices can have different id values when
1410	the broadcast packets are received only on one of them.
1411	The default value 0 means that the device is the only interface
1412	to its medium, value of -1 means that medium is not known.
1413
1414	Currently, it is used to change the proxy_arp behavior:
1415	the proxy_arp feature is enabled for packets forwarded between
1416	two devices attached to different media.
1417
1418proxy_arp - BOOLEAN
1419	Do proxy arp.
1420
1421	proxy_arp for the interface will be enabled if at least one of
1422	conf/{all,interface}/proxy_arp is set to TRUE,
1423	it will be disabled otherwise
1424
1425proxy_arp_pvlan - BOOLEAN
1426	Private VLAN proxy arp.
1427
1428	Basically allow proxy arp replies back to the same interface
1429	(from which the ARP request/solicitation was received).
1430
1431	This is done to support (ethernet) switch features, like RFC
1432	3069, where the individual ports are NOT allowed to
1433	communicate with each other, but they are allowed to talk to
1434	the upstream router.  As described in RFC 3069, it is possible
1435	to allow these hosts to communicate through the upstream
1436	router by proxy_arp'ing. Don't need to be used together with
1437	proxy_arp.
1438
1439	This technology is known by different names:
1440
1441	  In RFC 3069 it is called VLAN Aggregation.
1442	  Cisco and Allied Telesyn call it Private VLAN.
1443	  Hewlett-Packard call it Source-Port filtering or port-isolation.
1444	  Ericsson call it MAC-Forced Forwarding (RFC Draft).
1445
1446shared_media - BOOLEAN
1447	Send(router) or accept(host) RFC1620 shared media redirects.
1448	Overrides secure_redirects.
1449
1450	shared_media for the interface will be enabled if at least one of
1451	conf/{all,interface}/shared_media is set to TRUE,
1452	it will be disabled otherwise
1453
1454	default TRUE
1455
1456secure_redirects - BOOLEAN
1457	Accept ICMP redirect messages only to gateways listed in the
1458	interface's current gateway list. Even if disabled, RFC1122 redirect
1459	rules still apply.
1460
1461	Overridden by shared_media.
1462
1463	secure_redirects for the interface will be enabled if at least one of
1464	conf/{all,interface}/secure_redirects is set to TRUE,
1465	it will be disabled otherwise
1466
1467	default TRUE
1468
1469send_redirects - BOOLEAN
1470	Send redirects, if router.
1471
1472	send_redirects for the interface will be enabled if at least one of
1473	conf/{all,interface}/send_redirects is set to TRUE,
1474	it will be disabled otherwise
1475
1476	Default: TRUE
1477
1478bootp_relay - BOOLEAN
1479	Accept packets with source address 0.b.c.d destined
1480	not to this host as local ones. It is supposed, that
1481	BOOTP relay daemon will catch and forward such packets.
1482	conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
1483	for the interface
1484
1485	default FALSE
1486
1487	Not Implemented Yet.
1488
1489accept_source_route - BOOLEAN
1490	Accept packets with SRR option.
1491	conf/all/accept_source_route must also be set to TRUE to accept packets
1492	with SRR option on the interface
1493
1494	default
1495
1496		- TRUE (router)
1497		- FALSE (host)
1498
1499accept_local - BOOLEAN
1500	Accept packets with local source addresses. In combination with
1501	suitable routing, this can be used to direct packets between two
1502	local interfaces over the wire and have them accepted properly.
1503	default FALSE
1504
1505route_localnet - BOOLEAN
1506	Do not consider loopback addresses as martian source or destination
1507	while routing. This enables the use of 127/8 for local routing purposes.
1508
1509	default FALSE
1510
1511rp_filter - INTEGER
1512	- 0 - No source validation.
1513	- 1 - Strict mode as defined in RFC3704 Strict Reverse Path
1514	  Each incoming packet is tested against the FIB and if the interface
1515	  is not the best reverse path the packet check will fail.
1516	  By default failed packets are discarded.
1517	- 2 - Loose mode as defined in RFC3704 Loose Reverse Path
1518	  Each incoming packet's source address is also tested against the FIB
1519	  and if the source address is not reachable via any interface
1520	  the packet check will fail.
1521
1522	Current recommended practice in RFC3704 is to enable strict mode
1523	to prevent IP spoofing from DDos attacks. If using asymmetric routing
1524	or other complicated routing, then loose mode is recommended.
1525
1526	The max value from conf/{all,interface}/rp_filter is used
1527	when doing source validation on the {interface}.
1528
1529	Default value is 0. Note that some distributions enable it
1530	in startup scripts.
1531
1532src_valid_mark - BOOLEAN
1533	- 0 - The fwmark of the packet is not included in reverse path
1534	  route lookup.  This allows for asymmetric routing configurations
1535	  utilizing the fwmark in only one direction, e.g., transparent
1536	  proxying.
1537
1538	- 1 - The fwmark of the packet is included in reverse path route
1539	  lookup.  This permits rp_filter to function when the fwmark is
1540	  used for routing traffic in both directions.
1541
1542	This setting also affects the utilization of fmwark when
1543	performing source address selection for ICMP replies, or
1544	determining addresses stored for the IPOPT_TS_TSANDADDR and
1545	IPOPT_RR IP options.
1546
1547	The max value from conf/{all,interface}/src_valid_mark is used.
1548
1549	Default value is 0.
1550
1551arp_filter - BOOLEAN
1552	- 1 - Allows you to have multiple network interfaces on the same
1553	  subnet, and have the ARPs for each interface be answered
1554	  based on whether or not the kernel would route a packet from
1555	  the ARP'd IP out that interface (therefore you must use source
1556	  based routing for this to work). In other words it allows control
1557	  of which cards (usually 1) will respond to an arp request.
1558
1559	- 0 - (default) The kernel can respond to arp requests with addresses
1560	  from other interfaces. This may seem wrong but it usually makes
1561	  sense, because it increases the chance of successful communication.
1562	  IP addresses are owned by the complete host on Linux, not by
1563	  particular interfaces. Only for more complex setups like load-
1564	  balancing, does this behaviour cause problems.
1565
1566	arp_filter for the interface will be enabled if at least one of
1567	conf/{all,interface}/arp_filter is set to TRUE,
1568	it will be disabled otherwise
1569
1570arp_announce - INTEGER
1571	Define different restriction levels for announcing the local
1572	source IP address from IP packets in ARP requests sent on
1573	interface:
1574
1575	- 0 - (default) Use any local address, configured on any interface
1576	- 1 - Try to avoid local addresses that are not in the target's
1577	  subnet for this interface. This mode is useful when target
1578	  hosts reachable via this interface require the source IP
1579	  address in ARP requests to be part of their logical network
1580	  configured on the receiving interface. When we generate the
1581	  request we will check all our subnets that include the
1582	  target IP and will preserve the source address if it is from
1583	  such subnet. If there is no such subnet we select source
1584	  address according to the rules for level 2.
1585	- 2 - Always use the best local address for this target.
1586	  In this mode we ignore the source address in the IP packet
1587	  and try to select local address that we prefer for talks with
1588	  the target host. Such local address is selected by looking
1589	  for primary IP addresses on all our subnets on the outgoing
1590	  interface that include the target IP address. If no suitable
1591	  local address is found we select the first local address
1592	  we have on the outgoing interface or on all other interfaces,
1593	  with the hope we will receive reply for our request and
1594	  even sometimes no matter the source IP address we announce.
1595
1596	The max value from conf/{all,interface}/arp_announce is used.
1597
1598	Increasing the restriction level gives more chance for
1599	receiving answer from the resolved target while decreasing
1600	the level announces more valid sender's information.
1601
1602arp_ignore - INTEGER
1603	Define different modes for sending replies in response to
1604	received ARP requests that resolve local target IP addresses:
1605
1606	- 0 - (default): reply for any local target IP address, configured
1607	  on any interface
1608	- 1 - reply only if the target IP address is local address
1609	  configured on the incoming interface
1610	- 2 - reply only if the target IP address is local address
1611	  configured on the incoming interface and both with the
1612	  sender's IP address are part from same subnet on this interface
1613	- 3 - do not reply for local addresses configured with scope host,
1614	  only resolutions for global and link addresses are replied
1615	- 4-7 - reserved
1616	- 8 - do not reply for all local addresses
1617
1618	The max value from conf/{all,interface}/arp_ignore is used
1619	when ARP request is received on the {interface}
1620
1621arp_notify - BOOLEAN
1622	Define mode for notification of address and device changes.
1623
1624	 ==  ==========================================================
1625	  0  (default): do nothing
1626	  1  Generate gratuitous arp requests when device is brought up
1627	     or hardware address changes.
1628	 ==  ==========================================================
1629
1630arp_accept - BOOLEAN
1631	Define behavior for gratuitous ARP frames who's IP is not
1632	already present in the ARP table:
1633
1634	- 0 - don't create new entries in the ARP table
1635	- 1 - create new entries in the ARP table
1636
1637	Both replies and requests type gratuitous arp will trigger the
1638	ARP table to be updated, if this setting is on.
1639
1640	If the ARP table already contains the IP address of the
1641	gratuitous arp frame, the arp table will be updated regardless
1642	if this setting is on or off.
1643
1644arp_evict_nocarrier - BOOLEAN
1645	Clears the ARP cache on NOCARRIER events. This option is important for
1646	wireless devices where the ARP cache should not be cleared when roaming
1647	between access points on the same network. In most cases this should
1648	remain as the default (1).
1649
1650	- 1 - (default): Clear the ARP cache on NOCARRIER events
1651	- 0 - Do not clear ARP cache on NOCARRIER events
1652
1653mcast_solicit - INTEGER
1654	The maximum number of multicast probes in INCOMPLETE state,
1655	when the associated hardware address is unknown.  Defaults
1656	to 3.
1657
1658ucast_solicit - INTEGER
1659	The maximum number of unicast probes in PROBE state, when
1660	the hardware address is being reconfirmed.  Defaults to 3.
1661
1662app_solicit - INTEGER
1663	The maximum number of probes to send to the user space ARP daemon
1664	via netlink before dropping back to multicast probes (see
1665	mcast_resolicit).  Defaults to 0.
1666
1667mcast_resolicit - INTEGER
1668	The maximum number of multicast probes after unicast and
1669	app probes in PROBE state.  Defaults to 0.
1670
1671disable_policy - BOOLEAN
1672	Disable IPSEC policy (SPD) for this interface
1673
1674disable_xfrm - BOOLEAN
1675	Disable IPSEC encryption on this interface, whatever the policy
1676
1677igmpv2_unsolicited_report_interval - INTEGER
1678	The interval in milliseconds in which the next unsolicited
1679	IGMPv1 or IGMPv2 report retransmit will take place.
1680
1681	Default: 10000 (10 seconds)
1682
1683igmpv3_unsolicited_report_interval - INTEGER
1684	The interval in milliseconds in which the next unsolicited
1685	IGMPv3 report retransmit will take place.
1686
1687	Default: 1000 (1 seconds)
1688
1689ignore_routes_with_linkdown - BOOLEAN
1690        Ignore routes whose link is down when performing a FIB lookup.
1691
1692promote_secondaries - BOOLEAN
1693	When a primary IP address is removed from this interface
1694	promote a corresponding secondary IP address instead of
1695	removing all the corresponding secondary IP addresses.
1696
1697drop_unicast_in_l2_multicast - BOOLEAN
1698	Drop any unicast IP packets that are received in link-layer
1699	multicast (or broadcast) frames.
1700
1701	This behavior (for multicast) is actually a SHOULD in RFC
1702	1122, but is disabled by default for compatibility reasons.
1703
1704	Default: off (0)
1705
1706drop_gratuitous_arp - BOOLEAN
1707	Drop all gratuitous ARP frames, for example if there's a known
1708	good ARP proxy on the network and such frames need not be used
1709	(or in the case of 802.11, must not be used to prevent attacks.)
1710
1711	Default: off (0)
1712
1713
1714tag - INTEGER
1715	Allows you to write a number, which can be used as required.
1716
1717	Default value is 0.
1718
1719xfrm4_gc_thresh - INTEGER
1720	(Obsolete since linux-4.14)
1721	The threshold at which we will start garbage collecting for IPv4
1722	destination cache entries.  At twice this value the system will
1723	refuse new allocations.
1724
1725igmp_link_local_mcast_reports - BOOLEAN
1726	Enable IGMP reports for link local multicast groups in the
1727	224.0.0.X range.
1728
1729	Default TRUE
1730
1731Alexey Kuznetsov.
1732kuznet@ms2.inr.ac.ru
1733
1734Updated by:
1735
1736- Andi Kleen
1737  ak@muc.de
1738- Nicolas Delon
1739  delon.nicolas@wanadoo.fr
1740
1741
1742
1743
1744/proc/sys/net/ipv6/* Variables
1745==============================
1746
1747IPv6 has no global variables such as tcp_*.  tcp_* settings under ipv4/ also
1748apply to IPv6 [XXX?].
1749
1750bindv6only - BOOLEAN
1751	Default value for IPV6_V6ONLY socket option,
1752	which restricts use of the IPv6 socket to IPv6 communication
1753	only.
1754
1755		- TRUE: disable IPv4-mapped address feature
1756		- FALSE: enable IPv4-mapped address feature
1757
1758	Default: FALSE (as specified in RFC3493)
1759
1760flowlabel_consistency - BOOLEAN
1761	Protect the consistency (and unicity) of flow label.
1762	You have to disable it to use IPV6_FL_F_REFLECT flag on the
1763	flow label manager.
1764
1765	- TRUE: enabled
1766	- FALSE: disabled
1767
1768	Default: TRUE
1769
1770auto_flowlabels - INTEGER
1771	Automatically generate flow labels based on a flow hash of the
1772	packet. This allows intermediate devices, such as routers, to
1773	identify packet flows for mechanisms like Equal Cost Multipath
1774	Routing (see RFC 6438).
1775
1776	=  ===========================================================
1777	0  automatic flow labels are completely disabled
1778	1  automatic flow labels are enabled by default, they can be
1779	   disabled on a per socket basis using the IPV6_AUTOFLOWLABEL
1780	   socket option
1781	2  automatic flow labels are allowed, they may be enabled on a
1782	   per socket basis using the IPV6_AUTOFLOWLABEL socket option
1783	3  automatic flow labels are enabled and enforced, they cannot
1784	   be disabled by the socket option
1785	=  ===========================================================
1786
1787	Default: 1
1788
1789flowlabel_state_ranges - BOOLEAN
1790	Split the flow label number space into two ranges. 0-0x7FFFF is
1791	reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF
1792	is reserved for stateless flow labels as described in RFC6437.
1793
1794	- TRUE: enabled
1795	- FALSE: disabled
1796
1797	Default: true
1798
1799flowlabel_reflect - INTEGER
1800	Control flow label reflection. Needed for Path MTU
1801	Discovery to work with Equal Cost Multipath Routing in anycast
1802	environments. See RFC 7690 and:
1803	https://tools.ietf.org/html/draft-wang-6man-flow-label-reflection-01
1804
1805	This is a bitmask.
1806
1807	- 1: enabled for established flows
1808
1809	  Note that this prevents automatic flowlabel changes, as done
1810	  in "tcp: change IPv6 flow-label upon receiving spurious retransmission"
1811	  and "tcp: Change txhash on every SYN and RTO retransmit"
1812
1813	- 2: enabled for TCP RESET packets (no active listener)
1814	  If set, a RST packet sent in response to a SYN packet on a closed
1815	  port will reflect the incoming flow label.
1816
1817	- 4: enabled for ICMPv6 echo reply messages.
1818
1819	Default: 0
1820
1821fib_multipath_hash_policy - INTEGER
1822	Controls which hash policy to use for multipath routes.
1823
1824	Default: 0 (Layer 3)
1825
1826	Possible values:
1827
1828	- 0 - Layer 3 (source and destination addresses plus flow label)
1829	- 1 - Layer 4 (standard 5-tuple)
1830	- 2 - Layer 3 or inner Layer 3 if present
1831	- 3 - Custom multipath hash. Fields used for multipath hash calculation
1832	  are determined by fib_multipath_hash_fields sysctl
1833
1834fib_multipath_hash_fields - UNSIGNED INTEGER
1835	When fib_multipath_hash_policy is set to 3 (custom multipath hash), the
1836	fields used for multipath hash calculation are determined by this
1837	sysctl.
1838
1839	This value is a bitmask which enables various fields for multipath hash
1840	calculation.
1841
1842	Possible fields are:
1843
1844	====== ============================
1845	0x0001 Source IP address
1846	0x0002 Destination IP address
1847	0x0004 IP protocol
1848	0x0008 Flow Label
1849	0x0010 Source port
1850	0x0020 Destination port
1851	0x0040 Inner source IP address
1852	0x0080 Inner destination IP address
1853	0x0100 Inner IP protocol
1854	0x0200 Inner Flow Label
1855	0x0400 Inner source port
1856	0x0800 Inner destination port
1857	====== ============================
1858
1859	Default: 0x0007 (source IP, destination IP and IP protocol)
1860
1861anycast_src_echo_reply - BOOLEAN
1862	Controls the use of anycast addresses as source addresses for ICMPv6
1863	echo reply
1864
1865	- TRUE:  enabled
1866	- FALSE: disabled
1867
1868	Default: FALSE
1869
1870idgen_delay - INTEGER
1871	Controls the delay in seconds after which time to retry
1872	privacy stable address generation if a DAD conflict is
1873	detected.
1874
1875	Default: 1 (as specified in RFC7217)
1876
1877idgen_retries - INTEGER
1878	Controls the number of retries to generate a stable privacy
1879	address if a DAD conflict is detected.
1880
1881	Default: 3 (as specified in RFC7217)
1882
1883mld_qrv - INTEGER
1884	Controls the MLD query robustness variable (see RFC3810 9.1).
1885
1886	Default: 2 (as specified by RFC3810 9.1)
1887
1888	Minimum: 1 (as specified by RFC6636 4.5)
1889
1890max_dst_opts_number - INTEGER
1891	Maximum number of non-padding TLVs allowed in a Destination
1892	options extension header. If this value is less than zero
1893	then unknown options are disallowed and the number of known
1894	TLVs allowed is the absolute value of this number.
1895
1896	Default: 8
1897
1898max_hbh_opts_number - INTEGER
1899	Maximum number of non-padding TLVs allowed in a Hop-by-Hop
1900	options extension header. If this value is less than zero
1901	then unknown options are disallowed and the number of known
1902	TLVs allowed is the absolute value of this number.
1903
1904	Default: 8
1905
1906max_dst_opts_length - INTEGER
1907	Maximum length allowed for a Destination options extension
1908	header.
1909
1910	Default: INT_MAX (unlimited)
1911
1912max_hbh_length - INTEGER
1913	Maximum length allowed for a Hop-by-Hop options extension
1914	header.
1915
1916	Default: INT_MAX (unlimited)
1917
1918skip_notify_on_dev_down - BOOLEAN
1919	Controls whether an RTM_DELROUTE message is generated for routes
1920	removed when a device is taken down or deleted. IPv4 does not
1921	generate this message; IPv6 does by default. Setting this sysctl
1922	to true skips the message, making IPv4 and IPv6 on par in relying
1923	on userspace caches to track link events and evict routes.
1924
1925	Default: false (generate message)
1926
1927nexthop_compat_mode - BOOLEAN
1928	New nexthop API provides a means for managing nexthops independent of
1929	prefixes. Backwards compatibilty with old route format is enabled by
1930	default which means route dumps and notifications contain the new
1931	nexthop attribute but also the full, expanded nexthop definition.
1932	Further, updates or deletes of a nexthop configuration generate route
1933	notifications for each fib entry using the nexthop. Once a system
1934	understands the new API, this sysctl can be disabled to achieve full
1935	performance benefits of the new API by disabling the nexthop expansion
1936	and extraneous notifications.
1937	Default: true (backward compat mode)
1938
1939fib_notify_on_flag_change - INTEGER
1940        Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/
1941        RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed.
1942
1943        After installing a route to the kernel, user space receives an
1944        acknowledgment, which means the route was installed in the kernel,
1945        but not necessarily in hardware.
1946        It is also possible for a route already installed in hardware to change
1947        its action and therefore its flags. For example, a host route that is
1948        trapping packets can be "promoted" to perform decapsulation following
1949        the installation of an IPinIP/VXLAN tunnel.
1950        The notifications will indicate to user-space the state of the route.
1951
1952        Default: 0 (Do not emit notifications.)
1953
1954        Possible values:
1955
1956        - 0 - Do not emit notifications.
1957        - 1 - Emit notifications.
1958        - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change.
1959
1960ioam6_id - INTEGER
1961        Define the IOAM id of this node. Uses only 24 bits out of 32 in total.
1962
1963        Min: 0
1964        Max: 0xFFFFFF
1965
1966        Default: 0xFFFFFF
1967
1968ioam6_id_wide - LONG INTEGER
1969        Define the wide IOAM id of this node. Uses only 56 bits out of 64 in
1970        total. Can be different from ioam6_id.
1971
1972        Min: 0
1973        Max: 0xFFFFFFFFFFFFFF
1974
1975        Default: 0xFFFFFFFFFFFFFF
1976
1977IPv6 Fragmentation:
1978
1979ip6frag_high_thresh - INTEGER
1980	Maximum memory used to reassemble IPv6 fragments. When
1981	ip6frag_high_thresh bytes of memory is allocated for this purpose,
1982	the fragment handler will toss packets until ip6frag_low_thresh
1983	is reached.
1984
1985ip6frag_low_thresh - INTEGER
1986	See ip6frag_high_thresh
1987
1988ip6frag_time - INTEGER
1989	Time in seconds to keep an IPv6 fragment in memory.
1990
1991``conf/default/*``:
1992	Change the interface-specific default settings.
1993
1994	These settings would be used during creating new interfaces.
1995
1996
1997``conf/all/*``:
1998	Change all the interface-specific settings.
1999
2000	[XXX:  Other special features than forwarding?]
2001
2002conf/all/disable_ipv6 - BOOLEAN
2003	Changing this value is same as changing ``conf/default/disable_ipv6``
2004	setting and also all per-interface ``disable_ipv6`` settings to the same
2005	value.
2006
2007	Reading this value does not have any particular meaning. It does not say
2008	whether IPv6 support is enabled or disabled. Returned value can be 1
2009	also in the case when some interface has ``disable_ipv6`` set to 0 and
2010	has configured IPv6 addresses.
2011
2012conf/all/forwarding - BOOLEAN
2013	Enable global IPv6 forwarding between all interfaces.
2014
2015	IPv4 and IPv6 work differently here; e.g. netfilter must be used
2016	to control which interfaces may forward packets and which not.
2017
2018	This also sets all interfaces' Host/Router setting
2019	'forwarding' to the specified value.  See below for details.
2020
2021	This referred to as global forwarding.
2022
2023proxy_ndp - BOOLEAN
2024	Do proxy ndp.
2025
2026fwmark_reflect - BOOLEAN
2027	Controls the fwmark of kernel-generated IPv6 reply packets that are not
2028	associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
2029	If unset, these packets have a fwmark of zero. If set, they have the
2030	fwmark of the packet they are replying to.
2031
2032	Default: 0
2033
2034``conf/interface/*``:
2035	Change special settings per interface.
2036
2037	The functional behaviour for certain settings is different
2038	depending on whether local forwarding is enabled or not.
2039
2040accept_ra - INTEGER
2041	Accept Router Advertisements; autoconfigure using them.
2042
2043	It also determines whether or not to transmit Router
2044	Solicitations. If and only if the functional setting is to
2045	accept Router Advertisements, Router Solicitations will be
2046	transmitted.
2047
2048	Possible values are:
2049
2050		==  ===========================================================
2051		 0  Do not accept Router Advertisements.
2052		 1  Accept Router Advertisements if forwarding is disabled.
2053		 2  Overrule forwarding behaviour. Accept Router Advertisements
2054		    even if forwarding is enabled.
2055		==  ===========================================================
2056
2057	Functional default:
2058
2059		- enabled if local forwarding is disabled.
2060		- disabled if local forwarding is enabled.
2061
2062accept_ra_defrtr - BOOLEAN
2063	Learn default router in Router Advertisement.
2064
2065	Functional default:
2066
2067		- enabled if accept_ra is enabled.
2068		- disabled if accept_ra is disabled.
2069
2070ra_defrtr_metric - UNSIGNED INTEGER
2071	Route metric for default route learned in Router Advertisement. This value
2072	will be assigned as metric for the default route learned via IPv6 Router
2073	Advertisement. Takes affect only if accept_ra_defrtr is enabled.
2074
2075	Possible values:
2076		1 to 0xFFFFFFFF
2077
2078		Default: IP6_RT_PRIO_USER i.e. 1024.
2079
2080accept_ra_from_local - BOOLEAN
2081	Accept RA with source-address that is found on local machine
2082	if the RA is otherwise proper and able to be accepted.
2083
2084	Default is to NOT accept these as it may be an un-intended
2085	network loop.
2086
2087	Functional default:
2088
2089	   - enabled if accept_ra_from_local is enabled
2090	     on a specific interface.
2091	   - disabled if accept_ra_from_local is disabled
2092	     on a specific interface.
2093
2094accept_ra_min_hop_limit - INTEGER
2095	Minimum hop limit Information in Router Advertisement.
2096
2097	Hop limit Information in Router Advertisement less than this
2098	variable shall be ignored.
2099
2100	Default: 1
2101
2102accept_ra_pinfo - BOOLEAN
2103	Learn Prefix Information in Router Advertisement.
2104
2105	Functional default:
2106
2107		- enabled if accept_ra is enabled.
2108		- disabled if accept_ra is disabled.
2109
2110accept_ra_rt_info_min_plen - INTEGER
2111	Minimum prefix length of Route Information in RA.
2112
2113	Route Information w/ prefix smaller than this variable shall
2114	be ignored.
2115
2116	Functional default:
2117
2118		* 0 if accept_ra_rtr_pref is enabled.
2119		* -1 if accept_ra_rtr_pref is disabled.
2120
2121accept_ra_rt_info_max_plen - INTEGER
2122	Maximum prefix length of Route Information in RA.
2123
2124	Route Information w/ prefix larger than this variable shall
2125	be ignored.
2126
2127	Functional default:
2128
2129		* 0 if accept_ra_rtr_pref is enabled.
2130		* -1 if accept_ra_rtr_pref is disabled.
2131
2132accept_ra_rtr_pref - BOOLEAN
2133	Accept Router Preference in RA.
2134
2135	Functional default:
2136
2137		- enabled if accept_ra is enabled.
2138		- disabled if accept_ra is disabled.
2139
2140accept_ra_mtu - BOOLEAN
2141	Apply the MTU value specified in RA option 5 (RFC4861). If
2142	disabled, the MTU specified in the RA will be ignored.
2143
2144	Functional default:
2145
2146		- enabled if accept_ra is enabled.
2147		- disabled if accept_ra is disabled.
2148
2149accept_redirects - BOOLEAN
2150	Accept Redirects.
2151
2152	Functional default:
2153
2154		- enabled if local forwarding is disabled.
2155		- disabled if local forwarding is enabled.
2156
2157accept_source_route - INTEGER
2158	Accept source routing (routing extension header).
2159
2160	- >= 0: Accept only routing header type 2.
2161	- < 0: Do not accept routing header.
2162
2163	Default: 0
2164
2165autoconf - BOOLEAN
2166	Autoconfigure addresses using Prefix Information in Router
2167	Advertisements.
2168
2169	Functional default:
2170
2171		- enabled if accept_ra_pinfo is enabled.
2172		- disabled if accept_ra_pinfo is disabled.
2173
2174dad_transmits - INTEGER
2175	The amount of Duplicate Address Detection probes to send.
2176
2177	Default: 1
2178
2179forwarding - INTEGER
2180	Configure interface-specific Host/Router behaviour.
2181
2182	.. note::
2183
2184	   It is recommended to have the same setting on all
2185	   interfaces; mixed router/host scenarios are rather uncommon.
2186
2187	Possible values are:
2188
2189		- 0 Forwarding disabled
2190		- 1 Forwarding enabled
2191
2192	**FALSE (0)**:
2193
2194	By default, Host behaviour is assumed.  This means:
2195
2196	1. IsRouter flag is not set in Neighbour Advertisements.
2197	2. If accept_ra is TRUE (default), transmit Router
2198	   Solicitations.
2199	3. If accept_ra is TRUE (default), accept Router
2200	   Advertisements (and do autoconfiguration).
2201	4. If accept_redirects is TRUE (default), accept Redirects.
2202
2203	**TRUE (1)**:
2204
2205	If local forwarding is enabled, Router behaviour is assumed.
2206	This means exactly the reverse from the above:
2207
2208	1. IsRouter flag is set in Neighbour Advertisements.
2209	2. Router Solicitations are not sent unless accept_ra is 2.
2210	3. Router Advertisements are ignored unless accept_ra is 2.
2211	4. Redirects are ignored.
2212
2213	Default: 0 (disabled) if global forwarding is disabled (default),
2214	otherwise 1 (enabled).
2215
2216hop_limit - INTEGER
2217	Default Hop Limit to set.
2218
2219	Default: 64
2220
2221mtu - INTEGER
2222	Default Maximum Transfer Unit
2223
2224	Default: 1280 (IPv6 required minimum)
2225
2226ip_nonlocal_bind - BOOLEAN
2227	If set, allows processes to bind() to non-local IPv6 addresses,
2228	which can be quite useful - but may break some applications.
2229
2230	Default: 0
2231
2232router_probe_interval - INTEGER
2233	Minimum interval (in seconds) between Router Probing described
2234	in RFC4191.
2235
2236	Default: 60
2237
2238router_solicitation_delay - INTEGER
2239	Number of seconds to wait after interface is brought up
2240	before sending Router Solicitations.
2241
2242	Default: 1
2243
2244router_solicitation_interval - INTEGER
2245	Number of seconds to wait between Router Solicitations.
2246
2247	Default: 4
2248
2249router_solicitations - INTEGER
2250	Number of Router Solicitations to send until assuming no
2251	routers are present.
2252
2253	Default: 3
2254
2255use_oif_addrs_only - BOOLEAN
2256	When enabled, the candidate source addresses for destinations
2257	routed via this interface are restricted to the set of addresses
2258	configured on this interface (vis. RFC 6724, section 4).
2259
2260	Default: false
2261
2262use_tempaddr - INTEGER
2263	Preference for Privacy Extensions (RFC3041).
2264
2265	  * <= 0 : disable Privacy Extensions
2266	  * == 1 : enable Privacy Extensions, but prefer public
2267	    addresses over temporary addresses.
2268	  * >  1 : enable Privacy Extensions and prefer temporary
2269	    addresses over public addresses.
2270
2271	Default:
2272
2273		* 0 (for most devices)
2274		* -1 (for point-to-point devices and loopback devices)
2275
2276temp_valid_lft - INTEGER
2277	valid lifetime (in seconds) for temporary addresses.
2278
2279	Default: 172800 (2 days)
2280
2281temp_prefered_lft - INTEGER
2282	Preferred lifetime (in seconds) for temporary addresses.
2283
2284	Default: 86400 (1 day)
2285
2286keep_addr_on_down - INTEGER
2287	Keep all IPv6 addresses on an interface down event. If set static
2288	global addresses with no expiration time are not flushed.
2289
2290	*   >0 : enabled
2291	*    0 : system default
2292	*   <0 : disabled
2293
2294	Default: 0 (addresses are removed)
2295
2296max_desync_factor - INTEGER
2297	Maximum value for DESYNC_FACTOR, which is a random value
2298	that ensures that clients don't synchronize with each
2299	other and generate new addresses at exactly the same time.
2300	value is in seconds.
2301
2302	Default: 600
2303
2304regen_max_retry - INTEGER
2305	Number of attempts before give up attempting to generate
2306	valid temporary addresses.
2307
2308	Default: 5
2309
2310max_addresses - INTEGER
2311	Maximum number of autoconfigured addresses per interface.  Setting
2312	to zero disables the limitation.  It is not recommended to set this
2313	value too large (or to zero) because it would be an easy way to
2314	crash the kernel by allowing too many addresses to be created.
2315
2316	Default: 16
2317
2318disable_ipv6 - BOOLEAN
2319	Disable IPv6 operation.  If accept_dad is set to 2, this value
2320	will be dynamically set to TRUE if DAD fails for the link-local
2321	address.
2322
2323	Default: FALSE (enable IPv6 operation)
2324
2325	When this value is changed from 1 to 0 (IPv6 is being enabled),
2326	it will dynamically create a link-local address on the given
2327	interface and start Duplicate Address Detection, if necessary.
2328
2329	When this value is changed from 0 to 1 (IPv6 is being disabled),
2330	it will dynamically delete all addresses and routes on the given
2331	interface. From now on it will not possible to add addresses/routes
2332	to the selected interface.
2333
2334accept_dad - INTEGER
2335	Whether to accept DAD (Duplicate Address Detection).
2336
2337	 == ==============================================================
2338	  0  Disable DAD
2339	  1  Enable DAD (default)
2340	  2  Enable DAD, and disable IPv6 operation if MAC-based duplicate
2341	     link-local address has been found.
2342	 == ==============================================================
2343
2344	DAD operation and mode on a given interface will be selected according
2345	to the maximum value of conf/{all,interface}/accept_dad.
2346
2347force_tllao - BOOLEAN
2348	Enable sending the target link-layer address option even when
2349	responding to a unicast neighbor solicitation.
2350
2351	Default: FALSE
2352
2353	Quoting from RFC 2461, section 4.4, Target link-layer address:
2354
2355	"The option MUST be included for multicast solicitations in order to
2356	avoid infinite Neighbor Solicitation "recursion" when the peer node
2357	does not have a cache entry to return a Neighbor Advertisements
2358	message.  When responding to unicast solicitations, the option can be
2359	omitted since the sender of the solicitation has the correct link-
2360	layer address; otherwise it would not have be able to send the unicast
2361	solicitation in the first place. However, including the link-layer
2362	address in this case adds little overhead and eliminates a potential
2363	race condition where the sender deletes the cached link-layer address
2364	prior to receiving a response to a previous solicitation."
2365
2366ndisc_notify - BOOLEAN
2367	Define mode for notification of address and device changes.
2368
2369	* 0 - (default): do nothing
2370	* 1 - Generate unsolicited neighbour advertisements when device is brought
2371	  up or hardware address changes.
2372
2373ndisc_tclass - INTEGER
2374	The IPv6 Traffic Class to use by default when sending IPv6 Neighbor
2375	Discovery (Router Solicitation, Router Advertisement, Neighbor
2376	Solicitation, Neighbor Advertisement, Redirect) messages.
2377	These 8 bits can be interpreted as 6 high order bits holding the DSCP
2378	value and 2 low order bits representing ECN (which you probably want
2379	to leave cleared).
2380
2381	* 0 - (default)
2382
2383ndisc_evict_nocarrier - BOOLEAN
2384	Clears the neighbor discovery table on NOCARRIER events. This option is
2385	important for wireless devices where the neighbor discovery cache should
2386	not be cleared when roaming between access points on the same network.
2387	In most cases this should remain as the default (1).
2388
2389	- 1 - (default): Clear neighbor discover cache on NOCARRIER events.
2390	- 0 - Do not clear neighbor discovery cache on NOCARRIER events.
2391
2392mldv1_unsolicited_report_interval - INTEGER
2393	The interval in milliseconds in which the next unsolicited
2394	MLDv1 report retransmit will take place.
2395
2396	Default: 10000 (10 seconds)
2397
2398mldv2_unsolicited_report_interval - INTEGER
2399	The interval in milliseconds in which the next unsolicited
2400	MLDv2 report retransmit will take place.
2401
2402	Default: 1000 (1 second)
2403
2404force_mld_version - INTEGER
2405	* 0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
2406	* 1 - Enforce to use MLD version 1
2407	* 2 - Enforce to use MLD version 2
2408
2409suppress_frag_ndisc - INTEGER
2410	Control RFC 6980 (Security Implications of IPv6 Fragmentation
2411	with IPv6 Neighbor Discovery) behavior:
2412
2413	* 1 - (default) discard fragmented neighbor discovery packets
2414	* 0 - allow fragmented neighbor discovery packets
2415
2416optimistic_dad - BOOLEAN
2417	Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
2418
2419	* 0: disabled (default)
2420	* 1: enabled
2421
2422	Optimistic Duplicate Address Detection for the interface will be enabled
2423	if at least one of conf/{all,interface}/optimistic_dad is set to 1,
2424	it will be disabled otherwise.
2425
2426use_optimistic - BOOLEAN
2427	If enabled, do not classify optimistic addresses as deprecated during
2428	source address selection.  Preferred addresses will still be chosen
2429	before optimistic addresses, subject to other ranking in the source
2430	address selection algorithm.
2431
2432	* 0: disabled (default)
2433	* 1: enabled
2434
2435	This will be enabled if at least one of
2436	conf/{all,interface}/use_optimistic is set to 1, disabled otherwise.
2437
2438stable_secret - IPv6 address
2439	This IPv6 address will be used as a secret to generate IPv6
2440	addresses for link-local addresses and autoconfigured
2441	ones. All addresses generated after setting this secret will
2442	be stable privacy ones by default. This can be changed via the
2443	addrgenmode ip-link. conf/default/stable_secret is used as the
2444	secret for the namespace, the interface specific ones can
2445	overwrite that. Writes to conf/all/stable_secret are refused.
2446
2447	It is recommended to generate this secret during installation
2448	of a system and keep it stable after that.
2449
2450	By default the stable secret is unset.
2451
2452addr_gen_mode - INTEGER
2453	Defines how link-local and autoconf addresses are generated.
2454
2455	=  =================================================================
2456	0  generate address based on EUI64 (default)
2457	1  do no generate a link-local address, use EUI64 for addresses
2458	   generated from autoconf
2459	2  generate stable privacy addresses, using the secret from
2460	   stable_secret (RFC7217)
2461	3  generate stable privacy addresses, using a random secret if unset
2462	=  =================================================================
2463
2464drop_unicast_in_l2_multicast - BOOLEAN
2465	Drop any unicast IPv6 packets that are received in link-layer
2466	multicast (or broadcast) frames.
2467
2468	By default this is turned off.
2469
2470drop_unsolicited_na - BOOLEAN
2471	Drop all unsolicited neighbor advertisements, for example if there's
2472	a known good NA proxy on the network and such frames need not be used
2473	(or in the case of 802.11, must not be used to prevent attacks.)
2474
2475	By default this is turned off.
2476
2477accept_untracked_na - BOOLEAN
2478	Add a new neighbour cache entry in STALE state for routers on receiving a
2479	neighbour advertisement (either solicited or unsolicited) with target
2480	link-layer address option specified if no neighbour entry is already
2481	present for the advertised IPv6 address. Without this knob, NAs received
2482	for untracked addresses (absent in neighbour cache) are silently ignored.
2483
2484	This is as per router-side behaviour documented in RFC9131.
2485
2486	This has lower precedence than drop_unsolicited_na.
2487
2488	This will optimize the return path for the initial off-link communication
2489	that is initiated by a directly connected host, by ensuring that
2490	the first-hop router which turns on this setting doesn't have to
2491	buffer the initial return packets to do neighbour-solicitation.
2492	The prerequisite is that the host is configured to send
2493	unsolicited neighbour advertisements on interface bringup.
2494	This setting should be used in conjunction with the ndisc_notify setting
2495	on the host to satisfy this prerequisite.
2496
2497	By default this is turned off.
2498
2499enhanced_dad - BOOLEAN
2500	Include a nonce option in the IPv6 neighbor solicitation messages used for
2501	duplicate address detection per RFC7527. A received DAD NS will only signal
2502	a duplicate address if the nonce is different. This avoids any false
2503	detection of duplicates due to loopback of the NS messages that we send.
2504	The nonce option will be sent on an interface unless both of
2505	conf/{all,interface}/enhanced_dad are set to FALSE.
2506
2507	Default: TRUE
2508
2509``icmp/*``:
2510===========
2511
2512ratelimit - INTEGER
2513	Limit the maximal rates for sending ICMPv6 messages.
2514
2515	0 to disable any limiting,
2516	otherwise the minimal space between responses in milliseconds.
2517
2518	Default: 1000
2519
2520ratemask - list of comma separated ranges
2521	For ICMPv6 message types matching the ranges in the ratemask, limit
2522	the sending of the message according to ratelimit parameter.
2523
2524	The format used for both input and output is a comma separated
2525	list of ranges (e.g. "0-127,129" for ICMPv6 message type 0 to 127 and
2526	129). Writing to the file will clear all previous ranges of ICMPv6
2527	message types and update the current list with the input.
2528
2529	Refer to: https://www.iana.org/assignments/icmpv6-parameters/icmpv6-parameters.xhtml
2530	for numerical values of ICMPv6 message types, e.g. echo request is 128
2531	and echo reply is 129.
2532
2533	Default: 0-1,3-127 (rate limit ICMPv6 errors except Packet Too Big)
2534
2535echo_ignore_all - BOOLEAN
2536	If set non-zero, then the kernel will ignore all ICMP ECHO
2537	requests sent to it over the IPv6 protocol.
2538
2539	Default: 0
2540
2541echo_ignore_multicast - BOOLEAN
2542	If set non-zero, then the kernel will ignore all ICMP ECHO
2543	requests sent to it over the IPv6 protocol via multicast.
2544
2545	Default: 0
2546
2547echo_ignore_anycast - BOOLEAN
2548	If set non-zero, then the kernel will ignore all ICMP ECHO
2549	requests sent to it over the IPv6 protocol destined to anycast address.
2550
2551	Default: 0
2552
2553xfrm6_gc_thresh - INTEGER
2554	(Obsolete since linux-4.14)
2555	The threshold at which we will start garbage collecting for IPv6
2556	destination cache entries.  At twice this value the system will
2557	refuse new allocations.
2558
2559
2560IPv6 Update by:
2561Pekka Savola <pekkas@netcore.fi>
2562YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
2563
2564
2565/proc/sys/net/bridge/* Variables:
2566=================================
2567
2568bridge-nf-call-arptables - BOOLEAN
2569	- 1 : pass bridged ARP traffic to arptables' FORWARD chain.
2570	- 0 : disable this.
2571
2572	Default: 1
2573
2574bridge-nf-call-iptables - BOOLEAN
2575	- 1 : pass bridged IPv4 traffic to iptables' chains.
2576	- 0 : disable this.
2577
2578	Default: 1
2579
2580bridge-nf-call-ip6tables - BOOLEAN
2581	- 1 : pass bridged IPv6 traffic to ip6tables' chains.
2582	- 0 : disable this.
2583
2584	Default: 1
2585
2586bridge-nf-filter-vlan-tagged - BOOLEAN
2587	- 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
2588	- 0 : disable this.
2589
2590	Default: 0
2591
2592bridge-nf-filter-pppoe-tagged - BOOLEAN
2593	- 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
2594	- 0 : disable this.
2595
2596	Default: 0
2597
2598bridge-nf-pass-vlan-input-dev - BOOLEAN
2599	- 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
2600	  interface on the bridge and set the netfilter input device to the
2601	  vlan. This allows use of e.g. "iptables -i br0.1" and makes the
2602	  REDIRECT target work with vlan-on-top-of-bridge interfaces.  When no
2603	  matching vlan interface is found, or this switch is off, the input
2604	  device is set to the bridge interface.
2605
2606	- 0: disable bridge netfilter vlan interface lookup.
2607
2608	Default: 0
2609
2610``proc/sys/net/sctp/*`` Variables:
2611==================================
2612
2613addip_enable - BOOLEAN
2614	Enable or disable extension of  Dynamic Address Reconfiguration
2615	(ADD-IP) functionality specified in RFC5061.  This extension provides
2616	the ability to dynamically add and remove new addresses for the SCTP
2617	associations.
2618
2619	1: Enable extension.
2620
2621	0: Disable extension.
2622
2623	Default: 0
2624
2625pf_enable - INTEGER
2626	Enable or disable pf (pf is short for potentially failed) state. A value
2627	of pf_retrans > path_max_retrans also disables pf state. That is, one of
2628	both pf_enable and pf_retrans > path_max_retrans can disable pf state.
2629	Since pf_retrans and path_max_retrans can be changed by userspace
2630	application, sometimes user expects to disable pf state by the value of
2631	pf_retrans > path_max_retrans, but occasionally the value of pf_retrans
2632	or path_max_retrans is changed by the user application, this pf state is
2633	enabled. As such, it is necessary to add this to dynamically enable
2634	and disable pf state. See:
2635	https://datatracker.ietf.org/doc/draft-ietf-tsvwg-sctp-failover for
2636	details.
2637
2638	1: Enable pf.
2639
2640	0: Disable pf.
2641
2642	Default: 1
2643
2644pf_expose - INTEGER
2645	Unset or enable/disable pf (pf is short for potentially failed) state
2646	exposure.  Applications can control the exposure of the PF path state
2647	in the SCTP_PEER_ADDR_CHANGE event and the SCTP_GET_PEER_ADDR_INFO
2648	sockopt.   When it's unset, no SCTP_PEER_ADDR_CHANGE event with
2649	SCTP_ADDR_PF state will be sent and a SCTP_PF-state transport info
2650	can be got via SCTP_GET_PEER_ADDR_INFO sockopt;  When it's enabled,
2651	a SCTP_PEER_ADDR_CHANGE event will be sent for a transport becoming
2652	SCTP_PF state and a SCTP_PF-state transport info can be got via
2653	SCTP_GET_PEER_ADDR_INFO sockopt;  When it's diabled, no
2654	SCTP_PEER_ADDR_CHANGE event will be sent and it returns -EACCES when
2655	trying to get a SCTP_PF-state transport info via SCTP_GET_PEER_ADDR_INFO
2656	sockopt.
2657
2658	0: Unset pf state exposure, Compatible with old applications.
2659
2660	1: Disable pf state exposure.
2661
2662	2: Enable pf state exposure.
2663
2664	Default: 0
2665
2666addip_noauth_enable - BOOLEAN
2667	Dynamic Address Reconfiguration (ADD-IP) requires the use of
2668	authentication to protect the operations of adding or removing new
2669	addresses.  This requirement is mandated so that unauthorized hosts
2670	would not be able to hijack associations.  However, older
2671	implementations may not have implemented this requirement while
2672	allowing the ADD-IP extension.  For reasons of interoperability,
2673	we provide this variable to control the enforcement of the
2674	authentication requirement.
2675
2676	== ===============================================================
2677	1  Allow ADD-IP extension to be used without authentication.  This
2678	   should only be set in a closed environment for interoperability
2679	   with older implementations.
2680
2681	0  Enforce the authentication requirement
2682	== ===============================================================
2683
2684	Default: 0
2685
2686auth_enable - BOOLEAN
2687	Enable or disable Authenticated Chunks extension.  This extension
2688	provides the ability to send and receive authenticated chunks and is
2689	required for secure operation of Dynamic Address Reconfiguration
2690	(ADD-IP) extension.
2691
2692	- 1: Enable this extension.
2693	- 0: Disable this extension.
2694
2695	Default: 0
2696
2697prsctp_enable - BOOLEAN
2698	Enable or disable the Partial Reliability extension (RFC3758) which
2699	is used to notify peers that a given DATA should no longer be expected.
2700
2701	- 1: Enable extension
2702	- 0: Disable
2703
2704	Default: 1
2705
2706max_burst - INTEGER
2707	The limit of the number of new packets that can be initially sent.  It
2708	controls how bursty the generated traffic can be.
2709
2710	Default: 4
2711
2712association_max_retrans - INTEGER
2713	Set the maximum number for retransmissions that an association can
2714	attempt deciding that the remote end is unreachable.  If this value
2715	is exceeded, the association is terminated.
2716
2717	Default: 10
2718
2719max_init_retransmits - INTEGER
2720	The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
2721	that an association will attempt before declaring the destination
2722	unreachable and terminating.
2723
2724	Default: 8
2725
2726path_max_retrans - INTEGER
2727	The maximum number of retransmissions that will be attempted on a given
2728	path.  Once this threshold is exceeded, the path is considered
2729	unreachable, and new traffic will use a different path when the
2730	association is multihomed.
2731
2732	Default: 5
2733
2734pf_retrans - INTEGER
2735	The number of retransmissions that will be attempted on a given path
2736	before traffic is redirected to an alternate transport (should one
2737	exist).  Note this is distinct from path_max_retrans, as a path that
2738	passes the pf_retrans threshold can still be used.  Its only
2739	deprioritized when a transmission path is selected by the stack.  This
2740	setting is primarily used to enable fast failover mechanisms without
2741	having to reduce path_max_retrans to a very low value.  See:
2742	http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
2743	for details.  Note also that a value of pf_retrans > path_max_retrans
2744	disables this feature. Since both pf_retrans and path_max_retrans can
2745	be changed by userspace application, a variable pf_enable is used to
2746	disable pf state.
2747
2748	Default: 0
2749
2750ps_retrans - INTEGER
2751	Primary.Switchover.Max.Retrans (PSMR), it's a tunable parameter coming
2752	from section-5 "Primary Path Switchover" in rfc7829.  The primary path
2753	will be changed to another active path when the path error counter on
2754	the old primary path exceeds PSMR, so that "the SCTP sender is allowed
2755	to continue data transmission on a new working path even when the old
2756	primary destination address becomes active again".   Note this feature
2757	is disabled by initializing 'ps_retrans' per netns as 0xffff by default,
2758	and its value can't be less than 'pf_retrans' when changing by sysctl.
2759
2760	Default: 0xffff
2761
2762rto_initial - INTEGER
2763	The initial round trip timeout value in milliseconds that will be used
2764	in calculating round trip times.  This is the initial time interval
2765	for retransmissions.
2766
2767	Default: 3000
2768
2769rto_max - INTEGER
2770	The maximum value (in milliseconds) of the round trip timeout.  This
2771	is the largest time interval that can elapse between retransmissions.
2772
2773	Default: 60000
2774
2775rto_min - INTEGER
2776	The minimum value (in milliseconds) of the round trip timeout.  This
2777	is the smallest time interval the can elapse between retransmissions.
2778
2779	Default: 1000
2780
2781hb_interval - INTEGER
2782	The interval (in milliseconds) between HEARTBEAT chunks.  These chunks
2783	are sent at the specified interval on idle paths to probe the state of
2784	a given path between 2 associations.
2785
2786	Default: 30000
2787
2788sack_timeout - INTEGER
2789	The amount of time (in milliseconds) that the implementation will wait
2790	to send a SACK.
2791
2792	Default: 200
2793
2794valid_cookie_life - INTEGER
2795	The default lifetime of the SCTP cookie (in milliseconds).  The cookie
2796	is used during association establishment.
2797
2798	Default: 60000
2799
2800cookie_preserve_enable - BOOLEAN
2801	Enable or disable the ability to extend the lifetime of the SCTP cookie
2802	that is used during the establishment phase of SCTP association
2803
2804	- 1: Enable cookie lifetime extension.
2805	- 0: Disable
2806
2807	Default: 1
2808
2809cookie_hmac_alg - STRING
2810	Select the hmac algorithm used when generating the cookie value sent by
2811	a listening sctp socket to a connecting client in the INIT-ACK chunk.
2812	Valid values are:
2813
2814	* md5
2815	* sha1
2816	* none
2817
2818	Ability to assign md5 or sha1 as the selected alg is predicated on the
2819	configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and
2820	CONFIG_CRYPTO_SHA1).
2821
2822	Default: Dependent on configuration.  MD5 if available, else SHA1 if
2823	available, else none.
2824
2825rcvbuf_policy - INTEGER
2826	Determines if the receive buffer is attributed to the socket or to
2827	association.   SCTP supports the capability to create multiple
2828	associations on a single socket.  When using this capability, it is
2829	possible that a single stalled association that's buffering a lot
2830	of data may block other associations from delivering their data by
2831	consuming all of the receive buffer space.  To work around this,
2832	the rcvbuf_policy could be set to attribute the receiver buffer space
2833	to each association instead of the socket.  This prevents the described
2834	blocking.
2835
2836	- 1: rcvbuf space is per association
2837	- 0: rcvbuf space is per socket
2838
2839	Default: 0
2840
2841sndbuf_policy - INTEGER
2842	Similar to rcvbuf_policy above, this applies to send buffer space.
2843
2844	- 1: Send buffer is tracked per association
2845	- 0: Send buffer is tracked per socket.
2846
2847	Default: 0
2848
2849sctp_mem - vector of 3 INTEGERs: min, pressure, max
2850	Number of pages allowed for queueing by all SCTP sockets.
2851
2852	min: Below this number of pages SCTP is not bothered about its
2853	memory appetite. When amount of memory allocated by SCTP exceeds
2854	this number, SCTP starts to moderate memory usage.
2855
2856	pressure: This value was introduced to follow format of tcp_mem.
2857
2858	max: Number of pages allowed for queueing by all SCTP sockets.
2859
2860	Default is calculated at boot time from amount of available memory.
2861
2862sctp_rmem - vector of 3 INTEGERs: min, default, max
2863	Only the first value ("min") is used, "default" and "max" are
2864	ignored.
2865
2866	min: Minimal size of receive buffer used by SCTP socket.
2867	It is guaranteed to each SCTP socket (but not association) even
2868	under moderate memory pressure.
2869
2870	Default: 4K
2871
2872sctp_wmem  - vector of 3 INTEGERs: min, default, max
2873	Currently this tunable has no effect.
2874
2875addr_scope_policy - INTEGER
2876	Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
2877
2878	- 0   - Disable IPv4 address scoping
2879	- 1   - Enable IPv4 address scoping
2880	- 2   - Follow draft but allow IPv4 private addresses
2881	- 3   - Follow draft but allow IPv4 link local addresses
2882
2883	Default: 1
2884
2885udp_port - INTEGER
2886	The listening port for the local UDP tunneling sock. Normally it's
2887	using the IANA-assigned UDP port number 9899 (sctp-tunneling).
2888
2889	This UDP sock is used for processing the incoming UDP-encapsulated
2890	SCTP packets (from RFC6951), and shared by all applications in the
2891	same net namespace. This UDP sock will be closed when the value is
2892	set to 0.
2893
2894	The value will also be used to set the src port of the UDP header
2895	for the outgoing UDP-encapsulated SCTP packets. For the dest port,
2896	please refer to 'encap_port' below.
2897
2898	Default: 0
2899
2900encap_port - INTEGER
2901	The default remote UDP encapsulation port.
2902
2903	This value is used to set the dest port of the UDP header for the
2904	outgoing UDP-encapsulated SCTP packets by default. Users can also
2905	change the value for each sock/asoc/transport by using setsockopt.
2906	For further information, please refer to RFC6951.
2907
2908	Note that when connecting to a remote server, the client should set
2909	this to the port that the UDP tunneling sock on the peer server is
2910	listening to and the local UDP tunneling sock on the client also
2911	must be started. On the server, it would get the encap_port from
2912	the incoming packet's source port.
2913
2914	Default: 0
2915
2916plpmtud_probe_interval - INTEGER
2917        The time interval (in milliseconds) for the PLPMTUD probe timer,
2918        which is configured to expire after this period to receive an
2919        acknowledgment to a probe packet. This is also the time interval
2920        between the probes for the current pmtu when the probe search
2921        is done.
2922
2923        PLPMTUD will be disabled when 0 is set, and other values for it
2924        must be >= 5000.
2925
2926	Default: 0
2927
2928
2929``/proc/sys/net/core/*``
2930========================
2931
2932	Please see: Documentation/admin-guide/sysctl/net.rst for descriptions of these entries.
2933
2934
2935``/proc/sys/net/unix/*``
2936========================
2937
2938max_dgram_qlen - INTEGER
2939	The maximum length of dgram socket receive queue
2940
2941	Default: 10
2942
2943