xref: /linux/Documentation/networking/ip-sysctl.rst (revision 6770eaad75ae039b620e9dc05f62dcbcfc2e5085)
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	Forward Packets between interfaces.
12
13	This variable is special, its change resets all configuration
14	parameters to their default state (RFC1122 for hosts, RFC1812
15	for routers)
16
17	Possible values:
18
19	- 0 (disabled)
20	- 1 (enabled)
21
22	Default: 0 (disabled)
23
24ip_default_ttl - INTEGER
25	Default value of TTL field (Time To Live) for outgoing (but not
26	forwarded) IP packets. Should be between 1 and 255 inclusive.
27	Default: 64 (as recommended by RFC1700)
28
29ip_no_pmtu_disc - INTEGER
30	Disable Path MTU Discovery. If enabled in mode 1 and a
31	fragmentation-required ICMP is received, the PMTU to this
32	destination will be set to the smallest of the old MTU to
33	this destination and min_pmtu (see below). You will need
34	to raise min_pmtu to the smallest interface MTU on your system
35	manually if you want to avoid locally generated fragments.
36
37	In mode 2 incoming Path MTU Discovery messages will be
38	discarded. Outgoing frames are handled the same as in mode 1,
39	implicitly setting IP_PMTUDISC_DONT on every created socket.
40
41	Mode 3 is a hardened pmtu discover mode. The kernel will only
42	accept fragmentation-needed errors if the underlying protocol
43	can verify them besides a plain socket lookup. Current
44	protocols for which pmtu events will be honored are TCP and
45	SCTP as they verify e.g. the sequence number or the
46	association. This mode should not be enabled globally but is
47	only intended to secure e.g. name servers in namespaces where
48	TCP path mtu must still work but path MTU information of other
49	protocols should be discarded. If enabled globally this mode
50	could break other protocols.
51
52	Possible values: 0-3
53
54	Default: FALSE
55
56min_pmtu - INTEGER
57	default 552 - minimum Path MTU. Unless this is changed manually,
58	each cached pmtu will never be lower than this setting.
59
60ip_forward_use_pmtu - BOOLEAN
61	By default we don't trust protocol path MTUs while forwarding
62	because they could be easily forged and can lead to unwanted
63	fragmentation by the router.
64	You only need to enable this if you have user-space software
65	which tries to discover path mtus by itself and depends on the
66	kernel honoring this information. This is normally not the
67	case.
68
69	Possible values:
70
71	- 0 (disabled)
72	- 1 (enabled)
73
74	Default: 0 (disabled)
75
76fwmark_reflect - BOOLEAN
77	Controls the fwmark of kernel-generated IPv4 reply packets that are not
78	associated with a socket for example, TCP RSTs or ICMP echo replies).
79	If disabled, these packets have a fwmark of zero. If enabled, they have the
80	fwmark of the packet they are replying to.
81
82	Possible values:
83
84	- 0 (disabled)
85	- 1 (enabled)
86
87	Default: 0 (disabled)
88
89fib_multipath_use_neigh - BOOLEAN
90	Use status of existing neighbor entry when determining nexthop for
91	multipath routes. If disabled, neighbor information is not used and
92	packets could be directed to a failed nexthop. Only valid for kernels
93	built with CONFIG_IP_ROUTE_MULTIPATH enabled.
94
95	Possible values:
96
97	- 0 (disabled)
98	- 1 (enabled)
99
100	Default: 0 (disabled)
101
102fib_multipath_hash_policy - INTEGER
103	Controls which hash policy to use for multipath routes. Only valid
104	for kernels built with CONFIG_IP_ROUTE_MULTIPATH enabled.
105
106	Default: 0 (Layer 3)
107
108	Possible values:
109
110	- 0 - Layer 3
111	- 1 - Layer 4
112	- 2 - Layer 3 or inner Layer 3 if present
113	- 3 - Custom multipath hash. Fields used for multipath hash calculation
114	  are determined by fib_multipath_hash_fields sysctl
115
116fib_multipath_hash_fields - UNSIGNED INTEGER
117	When fib_multipath_hash_policy is set to 3 (custom multipath hash), the
118	fields used for multipath hash calculation are determined by this
119	sysctl.
120
121	This value is a bitmask which enables various fields for multipath hash
122	calculation.
123
124	Possible fields are:
125
126	====== ============================
127	0x0001 Source IP address
128	0x0002 Destination IP address
129	0x0004 IP protocol
130	0x0008 Unused (Flow Label)
131	0x0010 Source port
132	0x0020 Destination port
133	0x0040 Inner source IP address
134	0x0080 Inner destination IP address
135	0x0100 Inner IP protocol
136	0x0200 Inner Flow Label
137	0x0400 Inner source port
138	0x0800 Inner destination port
139	====== ============================
140
141	Default: 0x0007 (source IP, destination IP and IP protocol)
142
143fib_multipath_hash_seed - UNSIGNED INTEGER
144	The seed value used when calculating hash for multipath routes. Applies
145	to both IPv4 and IPv6 datapath. Only present for kernels built with
146	CONFIG_IP_ROUTE_MULTIPATH enabled.
147
148	When set to 0, the seed value used for multipath routing defaults to an
149	internal random-generated one.
150
151	The actual hashing algorithm is not specified -- there is no guarantee
152	that a next hop distribution effected by a given seed will keep stable
153	across kernel versions.
154
155	Default: 0 (random)
156
157fib_sync_mem - UNSIGNED INTEGER
158	Amount of dirty memory from fib entries that can be backlogged before
159	synchronize_rcu is forced.
160
161	Default: 512kB   Minimum: 64kB   Maximum: 64MB
162
163ip_forward_update_priority - INTEGER
164	Whether to update SKB priority from "TOS" field in IPv4 header after it
165	is forwarded. The new SKB priority is mapped from TOS field value
166	according to an rt_tos2priority table (see e.g. man tc-prio).
167
168	Default: 1 (Update priority.)
169
170	Possible values:
171
172	- 0 - Do not update priority.
173	- 1 - Update priority.
174
175route/max_size - INTEGER
176	Maximum number of routes allowed in the kernel.  Increase
177	this when using large numbers of interfaces and/or routes.
178
179	From linux kernel 3.6 onwards, this is deprecated for ipv4
180	as route cache is no longer used.
181
182	From linux kernel 6.3 onwards, this is deprecated for ipv6
183	as garbage collection manages cached route entries.
184
185neigh/default/gc_thresh1 - INTEGER
186	Minimum number of entries to keep.  Garbage collector will not
187	purge entries if there are fewer than this number.
188
189	Default: 128
190
191neigh/default/gc_thresh2 - INTEGER
192	Threshold when garbage collector becomes more aggressive about
193	purging entries. Entries older than 5 seconds will be cleared
194	when over this number.
195
196	Default: 512
197
198neigh/default/gc_thresh3 - INTEGER
199	Maximum number of non-PERMANENT neighbor entries allowed.  Increase
200	this when using large numbers of interfaces and when communicating
201	with large numbers of directly-connected peers.
202
203	Default: 1024
204
205neigh/default/unres_qlen_bytes - INTEGER
206	The maximum number of bytes which may be used by packets
207	queued for each	unresolved address by other network layers.
208	(added in linux 3.3)
209
210	Setting negative value is meaningless and will return error.
211
212	Default: SK_WMEM_DEFAULT, (same as net.core.wmem_default).
213
214		Exact value depends on architecture and kernel options,
215		but should be enough to allow queuing 256 packets
216		of medium size.
217
218neigh/default/unres_qlen - INTEGER
219	The maximum number of packets which may be queued for each
220	unresolved address by other network layers.
221
222	(deprecated in linux 3.3) : use unres_qlen_bytes instead.
223
224	Prior to linux 3.3, the default value is 3 which may cause
225	unexpected packet loss. The current default value is calculated
226	according to default value of unres_qlen_bytes and true size of
227	packet.
228
229	Default: 101
230
231neigh/default/interval_probe_time_ms - INTEGER
232	The probe interval for neighbor entries with NTF_MANAGED flag,
233	the min value is 1.
234
235	Default: 5000
236
237mtu_expires - INTEGER
238	Time, in seconds, that cached PMTU information is kept.
239
240min_adv_mss - INTEGER
241	The advertised MSS depends on the first hop route MTU, but will
242	never be lower than this setting.
243
244fib_notify_on_flag_change - INTEGER
245        Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/
246        RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed.
247
248        After installing a route to the kernel, user space receives an
249        acknowledgment, which means the route was installed in the kernel,
250        but not necessarily in hardware.
251        It is also possible for a route already installed in hardware to change
252        its action and therefore its flags. For example, a host route that is
253        trapping packets can be "promoted" to perform decapsulation following
254        the installation of an IPinIP/VXLAN tunnel.
255        The notifications will indicate to user-space the state of the route.
256
257        Default: 0 (Do not emit notifications.)
258
259        Possible values:
260
261        - 0 - Do not emit notifications.
262        - 1 - Emit notifications.
263        - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change.
264
265IP Fragmentation:
266
267ipfrag_high_thresh - LONG INTEGER
268	Maximum memory used to reassemble IP fragments.
269
270ipfrag_low_thresh - LONG INTEGER
271	(Obsolete since linux-4.17)
272	Maximum memory used to reassemble IP fragments before the kernel
273	begins to remove incomplete fragment queues to free up resources.
274	The kernel still accepts new fragments for defragmentation.
275
276ipfrag_time - INTEGER
277	Time in seconds to keep an IP fragment in memory.
278
279ipfrag_max_dist - INTEGER
280	ipfrag_max_dist is a non-negative integer value which defines the
281	maximum "disorder" which is allowed among fragments which share a
282	common IP source address. Note that reordering of packets is
283	not unusual, but if a large number of fragments arrive from a source
284	IP address while a particular fragment queue remains incomplete, it
285	probably indicates that one or more fragments belonging to that queue
286	have been lost. When ipfrag_max_dist is positive, an additional check
287	is done on fragments before they are added to a reassembly queue - if
288	ipfrag_max_dist (or more) fragments have arrived from a particular IP
289	address between additions to any IP fragment queue using that source
290	address, it's presumed that one or more fragments in the queue are
291	lost. The existing fragment queue will be dropped, and a new one
292	started. An ipfrag_max_dist value of zero disables this check.
293
294	Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
295	result in unnecessarily dropping fragment queues when normal
296	reordering of packets occurs, which could lead to poor application
297	performance. Using a very large value, e.g. 50000, increases the
298	likelihood of incorrectly reassembling IP fragments that originate
299	from different IP datagrams, which could result in data corruption.
300	Default: 64
301
302bc_forwarding - INTEGER
303	bc_forwarding enables the feature described in rfc1812#section-5.3.5.2
304	and rfc2644. It allows the router to forward directed broadcast.
305	To enable this feature, the 'all' entry and the input interface entry
306	should be set to 1.
307	Default: 0
308
309INET peer storage
310=================
311
312inet_peer_threshold - INTEGER
313	The approximate size of the storage.  Starting from this threshold
314	entries will be thrown aggressively.  This threshold also determines
315	entries' time-to-live and time intervals between garbage collection
316	passes.  More entries, less time-to-live, less GC interval.
317
318inet_peer_minttl - INTEGER
319	Minimum time-to-live of entries.  Should be enough to cover fragment
320	time-to-live on the reassembling side.  This minimum time-to-live  is
321	guaranteed if the pool size is less than inet_peer_threshold.
322	Measured in seconds.
323
324inet_peer_maxttl - INTEGER
325	Maximum time-to-live of entries.  Unused entries will expire after
326	this period of time if there is no memory pressure on the pool (i.e.
327	when the number of entries in the pool is very small).
328	Measured in seconds.
329
330TCP variables
331=============
332
333somaxconn - INTEGER
334	Limit of socket listen() backlog, known in userspace as SOMAXCONN.
335	Defaults to 4096. (Was 128 before linux-5.4)
336	See also tcp_max_syn_backlog for additional tuning for TCP sockets.
337
338tcp_abort_on_overflow - BOOLEAN
339	If listening service is too slow to accept new connections,
340	reset them. Default state is FALSE. It means that if overflow
341	occurred due to a burst, connection will recover. Enable this
342	option _only_ if you are really sure that listening daemon
343	cannot be tuned to accept connections faster. Enabling this
344	option can harm clients of your server.
345
346tcp_adv_win_scale - INTEGER
347	Obsolete since linux-6.6
348	Count buffering overhead as bytes/2^tcp_adv_win_scale
349	(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
350	if it is <= 0.
351
352	Possible values are [-31, 31], inclusive.
353
354	Default: 1
355
356tcp_allowed_congestion_control - STRING
357	Show/set the congestion control choices available to non-privileged
358	processes. The list is a subset of those listed in
359	tcp_available_congestion_control.
360
361	Default is "reno" and the default setting (tcp_congestion_control).
362
363tcp_app_win - INTEGER
364	Reserve max(window/2^tcp_app_win, mss) of window for application
365	buffer. Value 0 is special, it means that nothing is reserved.
366
367	Possible values are [0, 31], inclusive.
368
369	Default: 31
370
371tcp_autocorking - BOOLEAN
372	Enable TCP auto corking :
373	When applications do consecutive small write()/sendmsg() system calls,
374	we try to coalesce these small writes as much as possible, to lower
375	total amount of sent packets. This is done if at least one prior
376	packet for the flow is waiting in Qdisc queues or device transmit
377	queue. Applications can still use TCP_CORK for optimal behavior
378	when they know how/when to uncork their sockets.
379
380	Possible values:
381
382	- 0 (disabled)
383	- 1 (enabled)
384
385	Default: 1 (enabled)
386
387tcp_available_congestion_control - STRING
388	Shows the available congestion control choices that are registered.
389	More congestion control algorithms may be available as modules,
390	but not loaded.
391
392tcp_base_mss - INTEGER
393	The initial value of search_low to be used by the packetization layer
394	Path MTU discovery (MTU probing).  If MTU probing is enabled,
395	this is the initial MSS used by the connection.
396
397tcp_mtu_probe_floor - INTEGER
398	If MTU probing is enabled this caps the minimum MSS used for search_low
399	for the connection.
400
401	Default : 48
402
403tcp_min_snd_mss - INTEGER
404	TCP SYN and SYNACK messages usually advertise an ADVMSS option,
405	as described in RFC 1122 and RFC 6691.
406
407	If this ADVMSS option is smaller than tcp_min_snd_mss,
408	it is silently capped to tcp_min_snd_mss.
409
410	Default : 48 (at least 8 bytes of payload per segment)
411
412tcp_congestion_control - STRING
413	Set the congestion control algorithm to be used for new
414	connections. The algorithm "reno" is always available, but
415	additional choices may be available based on kernel configuration.
416	Default is set as part of kernel configuration.
417	For passive connections, the listener congestion control choice
418	is inherited.
419
420	[see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
421
422tcp_dsack - BOOLEAN
423	Allows TCP to send "duplicate" SACKs.
424
425	Possible values:
426
427	- 0 (disabled)
428	- 1 (enabled)
429
430	Default: 1 (enabled)
431
432tcp_early_retrans - INTEGER
433	Tail loss probe (TLP) converts RTOs occurring due to tail
434	losses into fast recovery (RFC8985). Note that
435	TLP requires RACK to function properly (see tcp_recovery below)
436
437	Possible values:
438
439		- 0 disables TLP
440		- 3 or 4 enables TLP
441
442	Default: 3
443
444tcp_ecn - INTEGER
445	Control use of Explicit Congestion Notification (ECN) by TCP.
446	ECN is used only when both ends of the TCP connection indicate support
447	for it. This feature is useful in avoiding losses due to congestion by
448	allowing supporting routers to signal congestion before having to drop
449	packets. A host that supports ECN both sends ECN at the IP layer and
450	feeds back ECN at the TCP layer. The highest variant of ECN feedback
451	that both peers support is chosen by the ECN negotiation (Accurate ECN,
452	ECN, or no ECN).
453
454	The highest negotiated variant for incoming connection requests
455	and the highest variant requested by outgoing connection
456	attempts:
457
458	===== ==================== ====================
459	Value Incoming connections Outgoing connections
460	===== ==================== ====================
461	0     No ECN               No ECN
462	1     ECN                  ECN
463	2     ECN                  No ECN
464	3     AccECN               AccECN
465	4     AccECN               ECN
466	5     AccECN               No ECN
467	===== ==================== ====================
468
469	Default: 2
470
471tcp_ecn_option - INTEGER
472	Control Accurate ECN (AccECN) option sending when AccECN has been
473	successfully negotiated during handshake. Send logic inhibits
474	sending AccECN options regarless of this setting when no AccECN
475	option has been seen for the reverse direction.
476
477	Possible values are:
478
479	= ============================================================
480	0 Never send AccECN option. This also disables sending AccECN
481	  option in SYN/ACK during handshake.
482	1 Send AccECN option sparingly according to the minimum option
483	  rules outlined in draft-ietf-tcpm-accurate-ecn.
484	2 Send AccECN option on every packet whenever it fits into TCP
485	  option space.
486	= ============================================================
487
488	Default: 2
489
490tcp_ecn_option_beacon - INTEGER
491	Control Accurate ECN (AccECN) option sending frequency per RTT and it
492	takes effect only when tcp_ecn_option is set to 2.
493
494	Default: 3 (AccECN will be send at least 3 times per RTT)
495
496tcp_ecn_fallback - BOOLEAN
497	If the kernel detects that ECN connection misbehaves, enable fall
498	back to non-ECN. Currently, this knob implements the fallback
499	from RFC3168, section 6.1.1.1., but we reserve that in future,
500	additional detection mechanisms could be implemented under this
501	knob. The value	is not used, if tcp_ecn or per route (or congestion
502	control) ECN settings are disabled.
503
504	Possible values:
505
506	- 0 (disabled)
507	- 1 (enabled)
508
509	Default: 1 (enabled)
510
511tcp_fack - BOOLEAN
512	This is a legacy option, it has no effect anymore.
513
514tcp_fin_timeout - INTEGER
515	The length of time an orphaned (no longer referenced by any
516	application) connection will remain in the FIN_WAIT_2 state
517	before it is aborted at the local end.  While a perfectly
518	valid "receive only" state for an un-orphaned connection, an
519	orphaned connection in FIN_WAIT_2 state could otherwise wait
520	forever for the remote to close its end of the connection.
521
522	Cf. tcp_max_orphans
523
524	Default: 60 seconds
525
526tcp_frto - INTEGER
527	Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
528	F-RTO is an enhanced recovery algorithm for TCP retransmission
529	timeouts.  It is particularly beneficial in networks where the
530	RTT fluctuates (e.g., wireless). F-RTO is sender-side only
531	modification. It does not require any support from the peer.
532
533	By default it's enabled with a non-zero value. 0 disables F-RTO.
534
535tcp_fwmark_accept - BOOLEAN
536	If enabled, incoming connections to listening sockets that do not have a
537	socket mark will set the mark of the accepting socket to the fwmark of
538	the incoming SYN packet. This will cause all packets on that connection
539	(starting from the first SYNACK) to be sent with that fwmark. The
540	listening socket's mark is unchanged. Listening sockets that already
541	have a fwmark set via setsockopt(SOL_SOCKET, SO_MARK, ...) are
542	unaffected.
543
544	Possible values:
545
546	- 0 (disabled)
547	- 1 (enabled)
548
549	Default: 0 (disabled)
550
551tcp_invalid_ratelimit - INTEGER
552	Limit the maximal rate for sending duplicate acknowledgments
553	in response to incoming TCP packets that are for an existing
554	connection but that are invalid due to any of these reasons:
555
556	  (a) out-of-window sequence number,
557	  (b) out-of-window acknowledgment number, or
558	  (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
559
560	This can help mitigate simple "ack loop" DoS attacks, wherein
561	a buggy or malicious middlebox or man-in-the-middle can
562	rewrite TCP header fields in manner that causes each endpoint
563	to think that the other is sending invalid TCP segments, thus
564	causing each side to send an unterminating stream of duplicate
565	acknowledgments for invalid segments.
566
567	Using 0 disables rate-limiting of dupacks in response to
568	invalid segments; otherwise this value specifies the minimal
569	space between sending such dupacks, in milliseconds.
570
571	Default: 500 (milliseconds).
572
573tcp_keepalive_time - INTEGER
574	How often TCP sends out keepalive messages when keepalive is enabled.
575	Default: 2hours.
576
577tcp_keepalive_probes - INTEGER
578	How many keepalive probes TCP sends out, until it decides that the
579	connection is broken. Default value: 9.
580
581tcp_keepalive_intvl - INTEGER
582	How frequently the probes are send out. Multiplied by
583	tcp_keepalive_probes it is time to kill not responding connection,
584	after probes started. Default value: 75sec i.e. connection
585	will be aborted after ~11 minutes of retries.
586
587tcp_l3mdev_accept - BOOLEAN
588	Enables child sockets to inherit the L3 master device index.
589	Enabling this option allows a "global" listen socket to work
590	across L3 master domains (e.g., VRFs) with connected sockets
591	derived from the listen socket to be bound to the L3 domain in
592	which the packets originated. Only valid when the kernel was
593	compiled with CONFIG_NET_L3_MASTER_DEV.
594
595	Possible values:
596
597	- 0 (disabled)
598	- 1 (enabled)
599
600	Default: 0 (disabled)
601
602tcp_low_latency - BOOLEAN
603	This is a legacy option, it has no effect anymore.
604
605tcp_max_orphans - INTEGER
606	Maximal number of TCP sockets not attached to any user file handle,
607	held by system.	If this number is exceeded orphaned connections are
608	reset immediately and warning is printed. This limit exists
609	only to prevent simple DoS attacks, you _must_ not rely on this
610	or lower the limit artificially, but rather increase it
611	(probably, after increasing installed memory),
612	if network conditions require more than default value,
613	and tune network services to linger and kill such states
614	more aggressively. Let me to remind again: each orphan eats
615	up to ~64K of unswappable memory.
616
617tcp_max_syn_backlog - INTEGER
618	Maximal number of remembered connection requests (SYN_RECV),
619	which have not received an acknowledgment from connecting client.
620
621	This is a per-listener limit.
622
623	The minimal value is 128 for low memory machines, and it will
624	increase in proportion to the memory of machine.
625
626	If server suffers from overload, try increasing this number.
627
628	Remember to also check /proc/sys/net/core/somaxconn
629	A SYN_RECV request socket consumes about 304 bytes of memory.
630
631tcp_max_tw_buckets - INTEGER
632	Maximal number of timewait sockets held by system simultaneously.
633	If this number is exceeded time-wait socket is immediately destroyed
634	and warning is printed. This limit exists only to prevent
635	simple DoS attacks, you _must_ not lower the limit artificially,
636	but rather increase it (probably, after increasing installed memory),
637	if network conditions require more than default value.
638
639tcp_mem - vector of 3 INTEGERs: min, pressure, max
640	min: below this number of pages TCP is not bothered about its
641	memory appetite.
642
643	pressure: when amount of memory allocated by TCP exceeds this number
644	of pages, TCP moderates its memory consumption and enters memory
645	pressure mode, which is exited when memory consumption falls
646	under "min".
647
648	max: number of pages allowed for queueing by all TCP sockets.
649
650	Defaults are calculated at boot time from amount of available
651	memory.
652
653tcp_min_rtt_wlen - INTEGER
654	The window length of the windowed min filter to track the minimum RTT.
655	A shorter window lets a flow more quickly pick up new (higher)
656	minimum RTT when it is moved to a longer path (e.g., due to traffic
657	engineering). A longer window makes the filter more resistant to RTT
658	inflations such as transient congestion. The unit is seconds.
659
660	Possible values: 0 - 86400 (1 day)
661
662	Default: 300
663
664tcp_moderate_rcvbuf - BOOLEAN
665	If enabled, TCP performs receive buffer auto-tuning, attempting to
666	automatically size the buffer (no greater than tcp_rmem[2]) to
667	match the size required by the path for full throughput.
668
669	Possible values:
670
671	- 0 (disabled)
672	- 1 (enabled)
673
674	Default: 1 (enabled)
675
676tcp_mtu_probing - INTEGER
677	Controls TCP Packetization-Layer Path MTU Discovery.  Takes three
678	values:
679
680	- 0 - Disabled
681	- 1 - Disabled by default, enabled when an ICMP black hole detected
682	- 2 - Always enabled, use initial MSS of tcp_base_mss.
683
684tcp_probe_interval - UNSIGNED INTEGER
685	Controls how often to start TCP Packetization-Layer Path MTU
686	Discovery reprobe. The default is reprobing every 10 minutes as
687	per RFC4821.
688
689tcp_probe_threshold - INTEGER
690	Controls when TCP Packetization-Layer Path MTU Discovery probing
691	will stop in respect to the width of search range in bytes. Default
692	is 8 bytes.
693
694tcp_no_metrics_save - BOOLEAN
695	By default, TCP saves various connection metrics in the route cache
696	when the connection closes, so that connections established in the
697	near future can use these to set initial conditions.  Usually, this
698	increases overall performance, but may sometimes cause performance
699	degradation.  If enabled, TCP will not cache metrics on closing
700	connections.
701
702	Possible values:
703
704	- 0 (disabled)
705	- 1 (enabled)
706
707	Default: 0 (disabled)
708
709tcp_no_ssthresh_metrics_save - BOOLEAN
710	Controls whether TCP saves ssthresh metrics in the route cache.
711	If enabled, ssthresh metrics are disabled.
712
713	Possible values:
714
715	- 0 (disabled)
716	- 1 (enabled)
717
718	Default: 1 (enabled)
719
720tcp_orphan_retries - INTEGER
721	This value influences the timeout of a locally closed TCP connection,
722	when RTO retransmissions remain unacknowledged.
723	See tcp_retries2 for more details.
724
725	The default value is 8.
726
727	If your machine is a loaded WEB server,
728	you should think about lowering this value, such sockets
729	may consume significant resources. Cf. tcp_max_orphans.
730
731tcp_recovery - INTEGER
732	This value is a bitmap to enable various experimental loss recovery
733	features.
734
735	=========   =============================================================
736	RACK: 0x1   enables RACK loss detection, for fast detection of lost
737		    retransmissions and tail drops, and resilience to
738		    reordering. currently, setting this bit to 0 has no
739		    effect, since RACK is the only supported loss detection
740		    algorithm.
741
742	RACK: 0x2   makes RACK's reordering window static (min_rtt/4).
743
744	RACK: 0x4   disables RACK's DUPACK threshold heuristic
745	=========   =============================================================
746
747	Default: 0x1
748
749tcp_reflect_tos - BOOLEAN
750	For listening sockets, reuse the DSCP value of the initial SYN message
751	for outgoing packets. This allows to have both directions of a TCP
752	stream to use the same DSCP value, assuming DSCP remains unchanged for
753	the lifetime of the connection.
754
755	This options affects both IPv4 and IPv6.
756
757	Possible values:
758
759	- 0 (disabled)
760	- 1 (enabled)
761
762	Default: 0 (disabled)
763
764tcp_reordering - INTEGER
765	Initial reordering level of packets in a TCP stream.
766	TCP stack can then dynamically adjust flow reordering level
767	between this initial value and tcp_max_reordering
768
769	Default: 3
770
771tcp_max_reordering - INTEGER
772	Maximal reordering level of packets in a TCP stream.
773	300 is a fairly conservative value, but you might increase it
774	if paths are using per packet load balancing (like bonding rr mode)
775
776	Default: 300
777
778tcp_retrans_collapse - BOOLEAN
779	Bug-to-bug compatibility with some broken printers.
780	On retransmit try to send bigger packets to work around bugs in
781	certain TCP stacks.
782
783	Possible values:
784
785	- 0 (disabled)
786	- 1 (enabled)
787
788	Default: 1 (enabled)
789
790tcp_retries1 - INTEGER
791	This value influences the time, after which TCP decides, that
792	something is wrong due to unacknowledged RTO retransmissions,
793	and reports this suspicion to the network layer.
794	See tcp_retries2 for more details.
795
796	RFC 1122 recommends at least 3 retransmissions, which is the
797	default.
798
799tcp_retries2 - INTEGER
800	This value influences the timeout of an alive TCP connection,
801	when RTO retransmissions remain unacknowledged.
802	Given a value of N, a hypothetical TCP connection following
803	exponential backoff with an initial RTO of TCP_RTO_MIN would
804	retransmit N times before killing the connection at the (N+1)th RTO.
805
806	The default value of 15 yields a hypothetical timeout of 924.6
807	seconds and is a lower bound for the effective timeout.
808	TCP will effectively time out at the first RTO which exceeds the
809	hypothetical timeout.
810	If tcp_rto_max_ms is decreased, it is recommended to also
811	change tcp_retries2.
812
813	RFC 1122 recommends at least 100 seconds for the timeout,
814	which corresponds to a value of at least 8.
815
816tcp_rfc1337 - BOOLEAN
817	If enabled, the TCP stack behaves conforming to RFC1337. If unset,
818	we are not conforming to RFC, but prevent TCP TIME_WAIT
819	assassination.
820
821	Possible values:
822
823	- 0 (disabled)
824	- 1 (enabled)
825
826	Default: 0 (disabled)
827
828tcp_rmem - vector of 3 INTEGERs: min, default, max
829	min: Minimal size of receive buffer used by TCP sockets.
830	It is guaranteed to each TCP socket, even under moderate memory
831	pressure.
832
833	Default: 4K
834
835	default: initial size of receive buffer used by TCP sockets.
836	This value overrides net.core.rmem_default used by other protocols.
837	Default: 131072 bytes.
838	This value results in initial window of 65535.
839
840	max: maximal size of receive buffer allowed for automatically
841	selected receiver buffers for TCP socket.
842	Calling setsockopt() with SO_RCVBUF disables
843	automatic tuning of that socket's receive buffer size, in which
844	case this value is ignored.
845	Default: between 131072 and 32MB, depending on RAM size.
846
847tcp_sack - BOOLEAN
848	Enable select acknowledgments (SACKS).
849
850	Possible values:
851
852	- 0 (disabled)
853	- 1 (enabled)
854
855	Default: 1 (enabled)
856
857tcp_comp_sack_rtt_percent - INTEGER
858	Percentage of SRTT used for the compressed SACK feature.
859	See tcp_comp_sack_nr, tcp_comp_sack_delay_ns, tcp_comp_sack_slack_ns.
860
861	Possible values : 1 - 1000
862
863	Default : 33 %
864
865tcp_comp_sack_delay_ns - LONG INTEGER
866	TCP tries to reduce number of SACK sent, using a timer based
867	on tcp_comp_sack_rtt_percent of SRTT, capped by this sysctl
868	in nano seconds.
869	The default is 1ms, based on TSO autosizing period.
870
871	Default : 1,000,000 ns (1 ms)
872
873tcp_comp_sack_slack_ns - LONG INTEGER
874	This sysctl control the slack used when arming the
875	timer used by SACK compression. This gives extra time
876	for small RTT flows, and reduces system overhead by allowing
877	opportunistic reduction of timer interrupts.
878	Too big values might reduce goodput.
879
880	Default : 10,000 ns (10 us)
881
882tcp_comp_sack_nr - INTEGER
883	Max number of SACK that can be compressed.
884	Using 0 disables SACK compression.
885
886	Default : 44
887
888tcp_backlog_ack_defer - BOOLEAN
889	If enabled, user thread processing socket backlog tries sending
890	one ACK for the whole queue. This helps to avoid potential
891	long latencies at end of a TCP socket syscall.
892
893	Possible values:
894
895	- 0 (disabled)
896	- 1 (enabled)
897
898	Default: 1 (enabled)
899
900tcp_slow_start_after_idle - BOOLEAN
901	If enabled, provide RFC2861 behavior and time out the congestion
902	window after an idle period.  An idle period is defined at
903	the current RTO.  If unset, the congestion window will not
904	be timed out after an idle period.
905
906	Possible values:
907
908	- 0 (disabled)
909	- 1 (enabled)
910
911	Default: 1 (enabled)
912
913tcp_stdurg - BOOLEAN
914	Use the Host requirements interpretation of the TCP urgent pointer field.
915	Most hosts use the older BSD interpretation, so if enabled,
916	Linux might not communicate correctly with them.
917
918	Possible values:
919
920	- 0 (disabled)
921	- 1 (enabled)
922
923	Default: 0 (disabled)
924
925tcp_synack_retries - INTEGER
926	Number of times SYNACKs for a passive TCP connection attempt will
927	be retransmitted. Should not be higher than 255. Default value
928	is 5, which corresponds to 31seconds till the last retransmission
929	with the current initial RTO of 1second. With this the final timeout
930	for a passive TCP connection will happen after 63seconds.
931
932tcp_syncookies - INTEGER
933	Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
934	Send out syncookies when the syn backlog queue of a socket
935	overflows. This is to prevent against the common 'SYN flood attack'
936	Default: 1
937
938	Note, that syncookies is fallback facility.
939	It MUST NOT be used to help highly loaded servers to stand
940	against legal connection rate. If you see SYN flood warnings
941	in your logs, but investigation	shows that they occur
942	because of overload with legal connections, you should tune
943	another parameters until this warning disappear.
944	See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
945
946	syncookies seriously violate TCP protocol, do not allow
947	to use TCP extensions, can result in serious degradation
948	of some services (f.e. SMTP relaying), visible not by you,
949	but your clients and relays, contacting you. While you see
950	SYN flood warnings in logs not being really flooded, your server
951	is seriously misconfigured.
952
953	If you want to test which effects syncookies have to your
954	network connections you can set this knob to 2 to enable
955	unconditionally generation of syncookies.
956
957tcp_migrate_req - BOOLEAN
958	The incoming connection is tied to a specific listening socket when
959	the initial SYN packet is received during the three-way handshake.
960	When a listener is closed, in-flight request sockets during the
961	handshake and established sockets in the accept queue are aborted.
962
963	If the listener has SO_REUSEPORT enabled, other listeners on the
964	same port should have been able to accept such connections. This
965	option makes it possible to migrate such child sockets to another
966	listener after close() or shutdown().
967
968	The BPF_SK_REUSEPORT_SELECT_OR_MIGRATE type of eBPF program should
969	usually be used to define the policy to pick an alive listener.
970	Otherwise, the kernel will randomly pick an alive listener only if
971	this option is enabled.
972
973	Note that migration between listeners with different settings may
974	crash applications. Let's say migration happens from listener A to
975	B, and only B has TCP_SAVE_SYN enabled. B cannot read SYN data from
976	the requests migrated from A. To avoid such a situation, cancel
977	migration by returning SK_DROP in the type of eBPF program, or
978	disable this option.
979
980	Possible values:
981
982	- 0 (disabled)
983	- 1 (enabled)
984
985	Default: 0 (disabled)
986
987tcp_fastopen - INTEGER
988	Enable TCP Fast Open (RFC7413) to send and accept data in the opening
989	SYN packet.
990
991	The client support is enabled by flag 0x1 (on by default). The client
992	then must use sendmsg() or sendto() with the MSG_FASTOPEN flag,
993	rather than connect() to send data in SYN.
994
995	The server support is enabled by flag 0x2 (off by default). Then
996	either enable for all listeners with another flag (0x400) or
997	enable individual listeners via TCP_FASTOPEN socket option with
998	the option value being the length of the syn-data backlog.
999
1000	The values (bitmap) are
1001
1002	=====  ======== ======================================================
1003	  0x1  (client) enables sending data in the opening SYN on the client.
1004	  0x2  (server) enables the server support, i.e., allowing data in
1005			a SYN packet to be accepted and passed to the
1006			application before 3-way handshake finishes.
1007	  0x4  (client) send data in the opening SYN regardless of cookie
1008			availability and without a cookie option.
1009	0x200  (server) accept data-in-SYN w/o any cookie option present.
1010	0x400  (server) enable all listeners to support Fast Open by
1011			default without explicit TCP_FASTOPEN socket option.
1012	=====  ======== ======================================================
1013
1014	Default: 0x1
1015
1016	Note that additional client or server features are only
1017	effective if the basic support (0x1 and 0x2) are enabled respectively.
1018
1019tcp_fastopen_blackhole_timeout_sec - INTEGER
1020	Initial time period in second to disable Fastopen on active TCP sockets
1021	when a TFO firewall blackhole issue happens.
1022	This time period will grow exponentially when more blackhole issues
1023	get detected right after Fastopen is re-enabled and will reset to
1024	initial value when the blackhole issue goes away.
1025	0 to disable the blackhole detection.
1026
1027	By default, it is set to 0 (feature is disabled).
1028
1029tcp_fastopen_key - list of comma separated 32-digit hexadecimal INTEGERs
1030	The list consists of a primary key and an optional backup key. The
1031	primary key is used for both creating and validating cookies, while the
1032	optional backup key is only used for validating cookies. The purpose of
1033	the backup key is to maximize TFO validation when keys are rotated.
1034
1035	A randomly chosen primary key may be configured by the kernel if
1036	the tcp_fastopen sysctl is set to 0x400 (see above), or if the
1037	TCP_FASTOPEN setsockopt() optname is set and a key has not been
1038	previously configured via sysctl. If keys are configured via
1039	setsockopt() by using the TCP_FASTOPEN_KEY optname, then those
1040	per-socket keys will be used instead of any keys that are specified via
1041	sysctl.
1042
1043	A key is specified as 4 8-digit hexadecimal integers which are separated
1044	by a '-' as: xxxxxxxx-xxxxxxxx-xxxxxxxx-xxxxxxxx. Leading zeros may be
1045	omitted. A primary and a backup key may be specified by separating them
1046	by a comma. If only one key is specified, it becomes the primary key and
1047	any previously configured backup keys are removed.
1048
1049tcp_syn_retries - INTEGER
1050	Number of times initial SYNs for an active TCP connection attempt
1051	will be retransmitted. Should not be higher than 127. Default value
1052	is 6, which corresponds to 67seconds (with tcp_syn_linear_timeouts = 4)
1053	till the last retransmission with the current initial RTO of 1second.
1054	With this the final timeout for an active TCP connection attempt
1055	will happen after 131seconds.
1056
1057tcp_timestamps - INTEGER
1058	Enable timestamps as defined in RFC1323.
1059
1060	- 0: Disabled.
1061	- 1: Enable timestamps as defined in RFC1323 and use random offset for
1062	  each connection rather than only using the current time.
1063	- 2: Like 1, but without random offsets.
1064
1065	Default: 1
1066
1067tcp_min_tso_segs - INTEGER
1068	Minimal number of segments per TSO frame.
1069
1070	Since linux-3.12, TCP does an automatic sizing of TSO frames,
1071	depending on flow rate, instead of filling 64Kbytes packets.
1072	For specific usages, it's possible to force TCP to build big
1073	TSO frames. Note that TCP stack might split too big TSO packets
1074	if available window is too small.
1075
1076	Default: 2
1077
1078tcp_tso_rtt_log - INTEGER
1079	Adjustment of TSO packet sizes based on min_rtt
1080
1081	Starting from linux-5.18, TCP autosizing can be tweaked
1082	for flows having small RTT.
1083
1084	Old autosizing was splitting the pacing budget to send 1024 TSO
1085	per second.
1086
1087	tso_packet_size = sk->sk_pacing_rate / 1024;
1088
1089	With the new mechanism, we increase this TSO sizing using:
1090
1091	distance = min_rtt_usec / (2^tcp_tso_rtt_log)
1092	tso_packet_size += gso_max_size >> distance;
1093
1094	This means that flows between very close hosts can use bigger
1095	TSO packets, reducing their cpu costs.
1096
1097	If you want to use the old autosizing, set this sysctl to 0.
1098
1099	Default: 9  (2^9 = 512 usec)
1100
1101tcp_pacing_ss_ratio - INTEGER
1102	sk->sk_pacing_rate is set by TCP stack using a ratio applied
1103	to current rate. (current_rate = cwnd * mss / srtt)
1104	If TCP is in slow start, tcp_pacing_ss_ratio is applied
1105	to let TCP probe for bigger speeds, assuming cwnd can be
1106	doubled every other RTT.
1107
1108	Default: 200
1109
1110tcp_pacing_ca_ratio - INTEGER
1111	sk->sk_pacing_rate is set by TCP stack using a ratio applied
1112	to current rate. (current_rate = cwnd * mss / srtt)
1113	If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio
1114	is applied to conservatively probe for bigger throughput.
1115
1116	Default: 120
1117
1118tcp_syn_linear_timeouts - INTEGER
1119	The number of times for an active TCP connection to retransmit SYNs with
1120	a linear backoff timeout before defaulting to an exponential backoff
1121	timeout. This has no effect on SYNACK at the passive TCP side.
1122
1123	With an initial RTO of 1 and tcp_syn_linear_timeouts = 4 we would
1124	expect SYN RTOs to be: 1, 1, 1, 1, 1, 2, 4, ... (4 linear timeouts,
1125	and the first exponential backoff using 2^0 * initial_RTO).
1126	Default: 4
1127
1128tcp_tso_win_divisor - INTEGER
1129	This allows control over what percentage of the congestion window
1130	can be consumed by a single TSO frame.
1131	The setting of this parameter is a choice between burstiness and
1132	building larger TSO frames.
1133
1134	Default: 3
1135
1136tcp_tw_reuse - INTEGER
1137	Enable reuse of TIME-WAIT sockets for new connections when it is
1138	safe from protocol viewpoint.
1139
1140	- 0 - disable
1141	- 1 - global enable
1142	- 2 - enable for loopback traffic only
1143
1144	It should not be changed without advice/request of technical
1145	experts.
1146
1147	Default: 2
1148
1149tcp_tw_reuse_delay - UNSIGNED INTEGER
1150        The delay in milliseconds before a TIME-WAIT socket can be reused by a
1151        new connection, if TIME-WAIT socket reuse is enabled. The actual reuse
1152        threshold is within [N, N+1] range, where N is the requested delay in
1153        milliseconds, to ensure the delay interval is never shorter than the
1154        configured value.
1155
1156        This setting contains an assumption about the other TCP timestamp clock
1157        tick interval. It should not be set to a value lower than the peer's
1158        clock tick for PAWS (Protection Against Wrapped Sequence numbers)
1159        mechanism work correctly for the reused connection.
1160
1161        Default: 1000 (milliseconds)
1162
1163tcp_window_scaling - BOOLEAN
1164	Enable window scaling as defined in RFC1323.
1165
1166	Possible values:
1167
1168	- 0 (disabled)
1169	- 1 (enabled)
1170
1171	Default: 1 (enabled)
1172
1173tcp_shrink_window - BOOLEAN
1174	This changes how the TCP receive window is calculated.
1175
1176	RFC 7323, section 2.4, says there are instances when a retracted
1177	window can be offered, and that TCP implementations MUST ensure
1178	that they handle a shrinking window, as specified in RFC 1122.
1179
1180	Possible values:
1181
1182	- 0 (disabled) - The window is never shrunk.
1183	- 1 (enabled)  - The window is shrunk when necessary to remain within
1184	  the memory limit set by autotuning (sk_rcvbuf).
1185	  This only occurs if a non-zero receive window
1186	  scaling factor is also in effect.
1187
1188	Default: 0 (disabled)
1189
1190tcp_wmem - vector of 3 INTEGERs: min, default, max
1191	min: Amount of memory reserved for send buffers for TCP sockets.
1192	Each TCP socket has rights to use it due to fact of its birth.
1193
1194	Default: 4K
1195
1196	default: initial size of send buffer used by TCP sockets.  This
1197	value overrides net.core.wmem_default used by other protocols.
1198
1199	It is usually lower than net.core.wmem_default.
1200
1201	Default: 16K
1202
1203	max: Maximal amount of memory allowed for automatically tuned
1204	send buffers for TCP sockets. This value does not override
1205	net.core.wmem_max.  Calling setsockopt() with SO_SNDBUF disables
1206	automatic tuning of that socket's send buffer size, in which case
1207	this value is ignored.
1208
1209	Default: between 64K and 4MB, depending on RAM size.
1210
1211tcp_notsent_lowat - UNSIGNED INTEGER
1212	A TCP socket can control the amount of unsent bytes in its write queue,
1213	thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
1214	reports POLLOUT events if the amount of unsent bytes is below a per
1215	socket value, and if the write queue is not full. sendmsg() will
1216	also not add new buffers if the limit is hit.
1217
1218	This global variable controls the amount of unsent data for
1219	sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
1220	to the global variable has immediate effect.
1221
1222	Default: UINT_MAX (0xFFFFFFFF)
1223
1224tcp_workaround_signed_windows - BOOLEAN
1225	If enabled, assume no receipt of a window scaling option means the
1226	remote TCP is broken and treats the window as a signed quantity.
1227	If disabled, assume the remote TCP is not broken even if we do
1228	not receive a window scaling option from them.
1229
1230	Possible values:
1231
1232	- 0 (disabled)
1233	- 1 (enabled)
1234
1235	Default: 0 (disabled)
1236
1237tcp_thin_linear_timeouts - BOOLEAN
1238	Enable dynamic triggering of linear timeouts for thin streams.
1239	If enabled, a check is performed upon retransmission by timeout to
1240	determine if the stream is thin (less than 4 packets in flight).
1241	As long as the stream is found to be thin, up to 6 linear
1242	timeouts may be performed before exponential backoff mode is
1243	initiated. This improves retransmission latency for
1244	non-aggressive thin streams, often found to be time-dependent.
1245	For more information on thin streams, see
1246	Documentation/networking/tcp-thin.rst
1247
1248	Possible values:
1249
1250	- 0 (disabled)
1251	- 1 (enabled)
1252
1253	Default: 0 (disabled)
1254
1255tcp_limit_output_bytes - INTEGER
1256	Controls TCP Small Queue limit per tcp socket.
1257	TCP bulk sender tends to increase packets in flight until it
1258	gets losses notifications. With SNDBUF autotuning, this can
1259	result in a large amount of packets queued on the local machine
1260	(e.g.: qdiscs, CPU backlog, or device) hurting latency of other
1261	flows, for typical pfifo_fast qdiscs.  tcp_limit_output_bytes
1262	limits the number of bytes on qdisc or device to reduce artificial
1263	RTT/cwnd and reduce bufferbloat.
1264
1265	Default: 4194304 (4 MB)
1266
1267tcp_challenge_ack_limit - INTEGER
1268	Limits number of Challenge ACK sent per second, as recommended
1269	in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
1270	Note that this per netns rate limit can allow some side channel
1271	attacks and probably should not be enabled.
1272	TCP stack implements per TCP socket limits anyway.
1273	Default: INT_MAX (unlimited)
1274
1275tcp_ehash_entries - INTEGER
1276	Show the number of hash buckets for TCP sockets in the current
1277	networking namespace.
1278
1279	A negative value means the networking namespace does not own its
1280	hash buckets and shares the initial networking namespace's one.
1281
1282tcp_child_ehash_entries - INTEGER
1283	Control the number of hash buckets for TCP sockets in the child
1284	networking namespace, which must be set before clone() or unshare().
1285
1286	If the value is not 0, the kernel uses a value rounded up to 2^n
1287	as the actual hash bucket size.  0 is a special value, meaning
1288	the child networking namespace will share the initial networking
1289	namespace's hash buckets.
1290
1291	Note that the child will use the global one in case the kernel
1292	fails to allocate enough memory.  In addition, the global hash
1293	buckets are spread over available NUMA nodes, but the allocation
1294	of the child hash table depends on the current process's NUMA
1295	policy, which could result in performance differences.
1296
1297	Note also that the default value of tcp_max_tw_buckets and
1298	tcp_max_syn_backlog depend on the hash bucket size.
1299
1300	Possible values: 0, 2^n (n: 0 - 24 (16Mi))
1301
1302	Default: 0
1303
1304tcp_plb_enabled - BOOLEAN
1305	If enabled and the underlying congestion control (e.g. DCTCP) supports
1306	and enables PLB feature, TCP PLB (Protective Load Balancing) is
1307	enabled. PLB is described in the following paper:
1308	https://doi.org/10.1145/3544216.3544226. Based on PLB parameters,
1309	upon sensing sustained congestion, TCP triggers a change in
1310	flow label field for outgoing IPv6 packets. A change in flow label
1311	field potentially changes the path of outgoing packets for switches
1312	that use ECMP/WCMP for routing.
1313
1314	PLB changes socket txhash which results in a change in IPv6 Flow Label
1315	field, and currently no-op for IPv4 headers. It is possible
1316	to apply PLB for IPv4 with other network header fields (e.g. TCP
1317	or IPv4 options) or using encapsulation where outer header is used
1318	by switches to determine next hop. In either case, further host
1319	and switch side changes will be needed.
1320
1321	If enabled, PLB assumes that congestion signal (e.g. ECN) is made
1322	available and used by congestion control module to estimate a
1323	congestion measure (e.g. ce_ratio). PLB needs a congestion measure to
1324	make repathing decisions.
1325
1326	Possible values:
1327
1328	- 0 (disabled)
1329	- 1 (enabled)
1330
1331	Default: 0 (disabled)
1332
1333tcp_plb_idle_rehash_rounds - INTEGER
1334	Number of consecutive congested rounds (RTT) seen after which
1335	a rehash can be performed, given there are no packets in flight.
1336	This is referred to as M in PLB paper:
1337	https://doi.org/10.1145/3544216.3544226.
1338
1339	Possible Values: 0 - 31
1340
1341	Default: 3
1342
1343tcp_plb_rehash_rounds - INTEGER
1344	Number of consecutive congested rounds (RTT) seen after which
1345	a forced rehash can be performed. Be careful when setting this
1346	parameter, as a small value increases the risk of retransmissions.
1347	This is referred to as N in PLB paper:
1348	https://doi.org/10.1145/3544216.3544226.
1349
1350	Possible Values: 0 - 31
1351
1352	Default: 12
1353
1354tcp_plb_suspend_rto_sec - INTEGER
1355	Time, in seconds, to suspend PLB in event of an RTO. In order to avoid
1356	having PLB repath onto a connectivity "black hole", after an RTO a TCP
1357	connection suspends PLB repathing for a random duration between 1x and
1358	2x of this parameter. Randomness is added to avoid concurrent rehashing
1359	of multiple TCP connections. This should be set corresponding to the
1360	amount of time it takes to repair a failed link.
1361
1362	Possible Values: 0 - 255
1363
1364	Default: 60
1365
1366tcp_plb_cong_thresh - INTEGER
1367	Fraction of packets marked with congestion over a round (RTT) to
1368	tag that round as congested. This is referred to as K in the PLB paper:
1369	https://doi.org/10.1145/3544216.3544226.
1370
1371	The 0-1 fraction range is mapped to 0-256 range to avoid floating
1372	point operations. For example, 128 means that if at least 50% of
1373	the packets in a round were marked as congested then the round
1374	will be tagged as congested.
1375
1376	Setting threshold to 0 means that PLB repaths every RTT regardless
1377	of congestion. This is not intended behavior for PLB and should be
1378	used only for experimentation purpose.
1379
1380	Possible Values: 0 - 256
1381
1382	Default: 128
1383
1384tcp_pingpong_thresh - INTEGER
1385	The number of estimated data replies sent for estimated incoming data
1386	requests that must happen before TCP considers that a connection is a
1387	"ping-pong" (request-response) connection for which delayed
1388	acknowledgments can provide benefits.
1389
1390	This threshold is 1 by default, but some applications may need a higher
1391	threshold for optimal performance.
1392
1393	Possible Values: 1 - 255
1394
1395	Default: 1
1396
1397tcp_rto_min_us - INTEGER
1398	Minimal TCP retransmission timeout (in microseconds). Note that the
1399	rto_min route option has the highest precedence for configuring this
1400	setting, followed by the TCP_BPF_RTO_MIN and TCP_RTO_MIN_US socket
1401	options, followed by this tcp_rto_min_us sysctl.
1402
1403	The recommended practice is to use a value less or equal to 200000
1404	microseconds.
1405
1406	Possible Values: 1 - INT_MAX
1407
1408	Default: 200000
1409
1410tcp_rto_max_ms - INTEGER
1411	Maximal TCP retransmission timeout (in ms).
1412	Note that TCP_RTO_MAX_MS socket option has higher precedence.
1413
1414	When changing tcp_rto_max_ms, it is important to understand
1415	that tcp_retries2 might need a change.
1416
1417	Possible Values: 1000 - 120,000
1418
1419	Default: 120,000
1420
1421UDP variables
1422=============
1423
1424udp_l3mdev_accept - BOOLEAN
1425	Enabling this option allows a "global" bound socket to work
1426	across L3 master domains (e.g., VRFs) with packets capable of
1427	being received regardless of the L3 domain in which they
1428	originated. Only valid when the kernel was compiled with
1429	CONFIG_NET_L3_MASTER_DEV.
1430
1431	Possible values:
1432
1433	- 0 (disabled)
1434	- 1 (enabled)
1435
1436	Default: 0 (disabled)
1437
1438udp_mem - vector of 3 INTEGERs: min, pressure, max
1439	Number of pages allowed for queueing by all UDP sockets.
1440
1441	min: Number of pages allowed for queueing by all UDP sockets.
1442
1443	pressure: This value was introduced to follow format of tcp_mem.
1444
1445	max: This value was introduced to follow format of tcp_mem.
1446
1447	Default is calculated at boot time from amount of available memory.
1448
1449udp_rmem_min - INTEGER
1450	Minimal size of receive buffer used by UDP sockets in moderation.
1451	Each UDP socket is able to use the size for receiving data, even if
1452	total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
1453
1454	Default: 4K
1455
1456udp_wmem_min - INTEGER
1457	UDP does not have tx memory accounting and this tunable has no effect.
1458
1459udp_hash_entries - INTEGER
1460	Show the number of hash buckets for UDP sockets in the current
1461	networking namespace.
1462
1463	A negative value means the networking namespace does not own its
1464	hash buckets and shares the initial networking namespace's one.
1465
1466udp_child_hash_entries - INTEGER
1467	Control the number of hash buckets for UDP sockets in the child
1468	networking namespace, which must be set before clone() or unshare().
1469
1470	If the value is not 0, the kernel uses a value rounded up to 2^n
1471	as the actual hash bucket size.  0 is a special value, meaning
1472	the child networking namespace will share the initial networking
1473	namespace's hash buckets.
1474
1475	Note that the child will use the global one in case the kernel
1476	fails to allocate enough memory.  In addition, the global hash
1477	buckets are spread over available NUMA nodes, but the allocation
1478	of the child hash table depends on the current process's NUMA
1479	policy, which could result in performance differences.
1480
1481	Possible values: 0, 2^n (n: 7 (128) - 16 (64K))
1482
1483	Default: 0
1484
1485
1486RAW variables
1487=============
1488
1489raw_l3mdev_accept - BOOLEAN
1490	Enabling this option allows a "global" bound socket to work
1491	across L3 master domains (e.g., VRFs) with packets capable of
1492	being received regardless of the L3 domain in which they
1493	originated. Only valid when the kernel was compiled with
1494	CONFIG_NET_L3_MASTER_DEV.
1495
1496	Possible values:
1497
1498	- 0 (disabled)
1499	- 1 (enabled)
1500
1501	Default: 1 (enabled)
1502
1503CIPSOv4 Variables
1504=================
1505
1506cipso_cache_enable - BOOLEAN
1507	If enabled, enable additions to and lookups from the CIPSO label mapping
1508	cache.  If disabled, additions are ignored and lookups always result in a
1509	miss.  However, regardless of the setting the cache is still
1510	invalidated when required when means you can safely toggle this on and
1511	off and the cache will always be "safe".
1512
1513	Possible values:
1514
1515	- 0 (disabled)
1516	- 1 (enabled)
1517
1518	Default: 1 (enabled)
1519
1520cipso_cache_bucket_size - INTEGER
1521	The CIPSO label cache consists of a fixed size hash table with each
1522	hash bucket containing a number of cache entries.  This variable limits
1523	the number of entries in each hash bucket; the larger the value is, the
1524	more CIPSO label mappings that can be cached.  When the number of
1525	entries in a given hash bucket reaches this limit adding new entries
1526	causes the oldest entry in the bucket to be removed to make room.
1527
1528	Default: 10
1529
1530cipso_rbm_optfmt - BOOLEAN
1531	Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
1532	the CIPSO draft specification (see Documentation/netlabel for details).
1533	This means that when set the CIPSO tag will be padded with empty
1534	categories in order to make the packet data 32-bit aligned.
1535
1536	Possible values:
1537
1538	- 0 (disabled)
1539	- 1 (enabled)
1540
1541	Default: 0 (disabled)
1542
1543cipso_rbm_strictvalid - BOOLEAN
1544	If enabled, do a very strict check of the CIPSO option when
1545	ip_options_compile() is called.  If disabled, relax the checks done during
1546	ip_options_compile().  Either way is "safe" as errors are caught else
1547	where in the CIPSO processing code but setting this to 0 (False) should
1548	result in less work (i.e. it should be faster) but could cause problems
1549	with other implementations that require strict checking.
1550
1551	Possible values:
1552
1553	- 0 (disabled)
1554	- 1 (enabled)
1555
1556	Default: 0 (disabled)
1557
1558IP Variables
1559============
1560
1561ip_local_port_range - 2 INTEGERS
1562	Defines the local port range that is used by TCP and UDP to
1563	choose the local port. The first number is the first, the
1564	second the last local port number.
1565	If possible, it is better these numbers have different parity
1566	(one even and one odd value).
1567	Must be greater than or equal to ip_unprivileged_port_start.
1568	The default values are 32768 and 60999 respectively.
1569
1570ip_local_reserved_ports - list of comma separated ranges
1571	Specify the ports which are reserved for known third-party
1572	applications. These ports will not be used by automatic port
1573	assignments (e.g. when calling connect() or bind() with port
1574	number 0). Explicit port allocation behavior is unchanged.
1575
1576	The format used for both input and output is a comma separated
1577	list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
1578	10). Writing to the file will clear all previously reserved
1579	ports and update the current list with the one given in the
1580	input.
1581
1582	Note that ip_local_port_range and ip_local_reserved_ports
1583	settings are independent and both are considered by the kernel
1584	when determining which ports are available for automatic port
1585	assignments.
1586
1587	You can reserve ports which are not in the current
1588	ip_local_port_range, e.g.::
1589
1590	    $ cat /proc/sys/net/ipv4/ip_local_port_range
1591	    32000	60999
1592	    $ cat /proc/sys/net/ipv4/ip_local_reserved_ports
1593	    8080,9148
1594
1595	although this is redundant. However such a setting is useful
1596	if later the port range is changed to a value that will
1597	include the reserved ports. Also keep in mind, that overlapping
1598	of these ranges may affect probability of selecting ephemeral
1599	ports which are right after block of reserved ports.
1600
1601	Default: Empty
1602
1603ip_unprivileged_port_start - INTEGER
1604	This is a per-namespace sysctl.  It defines the first
1605	unprivileged port in the network namespace.  Privileged ports
1606	require root or CAP_NET_BIND_SERVICE in order to bind to them.
1607	To disable all privileged ports, set this to 0.  They must not
1608	overlap with the ip_local_port_range.
1609
1610	Default: 1024
1611
1612ip_nonlocal_bind - BOOLEAN
1613	If enabled, allows processes to bind() to non-local IP addresses,
1614	which can be quite useful - but may break some applications.
1615
1616	Possible values:
1617
1618	- 0 (disabled)
1619	- 1 (enabled)
1620
1621	Default: 0 (disabled)
1622
1623ip_autobind_reuse - BOOLEAN
1624	By default, bind() does not select the ports automatically even if
1625	the new socket and all sockets bound to the port have SO_REUSEADDR.
1626	ip_autobind_reuse allows bind() to reuse the port and this is useful
1627	when you use bind()+connect(), but may break some applications.
1628	The preferred solution is to use IP_BIND_ADDRESS_NO_PORT and this
1629	option should only be set by experts.
1630
1631	Possible values:
1632
1633	- 0 (disabled)
1634	- 1 (enabled)
1635
1636	Default: 0 (disabled)
1637
1638ip_dynaddr - INTEGER
1639	If set non-zero, enables support for dynamic addresses.
1640	If set to a non-zero value larger than 1, a kernel log
1641	message will be printed when dynamic address rewriting
1642	occurs.
1643
1644	Default: 0
1645
1646ip_early_demux - BOOLEAN
1647	Optimize input packet processing down to one demux for
1648	certain kinds of local sockets.  Currently we only do this
1649	for established TCP and connected UDP sockets.
1650
1651	It may add an additional cost for pure routing workloads that
1652	reduces overall throughput, in such case you should disable it.
1653
1654	Possible values:
1655
1656	- 0 (disabled)
1657	- 1 (enabled)
1658
1659	Default: 1 (enabled)
1660
1661ping_group_range - 2 INTEGERS
1662	Restrict ICMP_PROTO datagram sockets to users in the group range.
1663	The default is "1 0", meaning, that nobody (not even root) may
1664	create ping sockets.  Setting it to "100 100" would grant permissions
1665	to the single group. "0 4294967294" would enable it for the world, "100
1666	4294967294" would enable it for the users, but not daemons.
1667
1668tcp_early_demux - BOOLEAN
1669	Enable early demux for established TCP sockets.
1670
1671	Possible values:
1672
1673	- 0 (disabled)
1674	- 1 (enabled)
1675
1676	Default: 1 (enabled)
1677
1678udp_early_demux - BOOLEAN
1679	Enable early demux for connected UDP sockets. Disable this if
1680	your system could experience more unconnected load.
1681
1682	Possible values:
1683
1684	- 0 (disabled)
1685	- 1 (enabled)
1686
1687	Default: 1 (enabled)
1688
1689icmp_echo_ignore_all - BOOLEAN
1690	If enabled, then the kernel will ignore all ICMP ECHO
1691	requests sent to it.
1692
1693	Possible values:
1694
1695	- 0 (disabled)
1696	- 1 (enabled)
1697
1698	Default: 0 (disabled)
1699
1700icmp_echo_enable_probe - BOOLEAN
1701        If enabled, then the kernel will respond to RFC 8335 PROBE
1702        requests sent to it.
1703
1704        Possible values:
1705
1706	- 0 (disabled)
1707	- 1 (enabled)
1708
1709	Default: 0 (disabled)
1710
1711icmp_echo_ignore_broadcasts - BOOLEAN
1712	If enabled, then the kernel will ignore all ICMP ECHO and
1713	TIMESTAMP requests sent to it via broadcast/multicast.
1714
1715	Possible values:
1716
1717	- 0 (disabled)
1718	- 1 (enabled)
1719
1720	Default: 1 (enabled)
1721
1722icmp_ratelimit - INTEGER
1723	Limit the maximal rates for sending ICMP packets whose type matches
1724	icmp_ratemask (see below) to specific targets.
1725	0 to disable any limiting,
1726	otherwise the minimal space between responses in milliseconds.
1727	Note that another sysctl, icmp_msgs_per_sec limits the number
1728	of ICMP packets	sent on all targets.
1729
1730	Default: 1000
1731
1732icmp_msgs_per_sec - INTEGER
1733	Limit maximal number of ICMP packets sent per second from this host.
1734	Only messages whose type matches icmp_ratemask (see below) are
1735	controlled by this limit. For security reasons, the precise count
1736	of messages per second is randomized.
1737
1738	Default: 1000
1739
1740icmp_msgs_burst - INTEGER
1741	icmp_msgs_per_sec controls number of ICMP packets sent per second,
1742	while icmp_msgs_burst controls the burst size of these packets.
1743	For security reasons, the precise burst size is randomized.
1744
1745	Default: 50
1746
1747icmp_ratemask - INTEGER
1748	Mask made of ICMP types for which rates are being limited.
1749
1750	Significant bits: IHGFEDCBA9876543210
1751
1752	Default mask:     0000001100000011000 (6168)
1753
1754	Bit definitions (see include/linux/icmp.h):
1755
1756		= =========================
1757		0 Echo Reply
1758		3 Destination Unreachable [1]_
1759		4 Source Quench [1]_
1760		5 Redirect
1761		8 Echo Request
1762		B Time Exceeded [1]_
1763		C Parameter Problem [1]_
1764		D Timestamp Request
1765		E Timestamp Reply
1766		F Info Request
1767		G Info Reply
1768		H Address Mask Request
1769		I Address Mask Reply
1770		= =========================
1771
1772	.. [1] These are rate limited by default (see default mask above)
1773
1774icmp_ignore_bogus_error_responses - BOOLEAN
1775	Some routers violate RFC1122 by sending bogus responses to broadcast
1776	frames.  Such violations are normally logged via a kernel warning.
1777	If enabled, the kernel will not give such warnings, which
1778	will avoid log file clutter.
1779
1780	Possible values:
1781
1782	- 0 (disabled)
1783	- 1 (enabled)
1784
1785	Default: 1 (enabled)
1786
1787icmp_errors_use_inbound_ifaddr - BOOLEAN
1788
1789	If disabled, icmp error messages are sent with the primary address of
1790	the exiting interface.
1791
1792	If enabled, the message will be sent with the primary address of
1793	the interface that received the packet that caused the icmp error.
1794	This is the behaviour many network administrators will expect from
1795	a router. And it can make debugging complicated network layouts
1796	much easier.
1797
1798	Note that if no primary address exists for the interface selected,
1799	then the primary address of the first non-loopback interface that
1800	has one will be used regardless of this setting.
1801
1802	Possible values:
1803
1804	- 0 (disabled)
1805	- 1 (enabled)
1806
1807	Default: 0 (disabled)
1808
1809icmp_errors_extension_mask - UNSIGNED INTEGER
1810	Bitmask of ICMP extensions to append to ICMPv4 error messages
1811	("Destination Unreachable", "Time Exceeded" and "Parameter Problem").
1812	The original datagram is trimmed / padded to 128 bytes in order to be
1813	compatible with applications that do not comply with RFC 4884.
1814
1815	Possible extensions are:
1816
1817	==== ==============================================================
1818	0x01 Incoming IP interface information according to RFC 5837.
1819	     Extension will include the index, IPv4 address (if present),
1820	     name and MTU of the IP interface that received the datagram
1821	     which elicited the ICMP error.
1822	==== ==============================================================
1823
1824	Default: 0x00 (no extensions)
1825
1826igmp_max_memberships - INTEGER
1827	Change the maximum number of multicast groups we can subscribe to.
1828	Default: 20
1829
1830	Theoretical maximum value is bounded by having to send a membership
1831	report in a single datagram (i.e. the report can't span multiple
1832	datagrams, or risk confusing the switch and leaving groups you don't
1833	intend to).
1834
1835	The number of supported groups 'M' is bounded by the number of group
1836	report entries you can fit into a single datagram of 65535 bytes.
1837
1838	M = 65536-sizeof (ip header)/(sizeof(Group record))
1839
1840	Group records are variable length, with a minimum of 12 bytes.
1841	So net.ipv4.igmp_max_memberships should not be set higher than:
1842
1843	(65536-24) / 12 = 5459
1844
1845	The value 5459 assumes no IP header options, so in practice
1846	this number may be lower.
1847
1848igmp_max_msf - INTEGER
1849	Maximum number of addresses allowed in the source filter list for a
1850	multicast group.
1851
1852	Default: 10
1853
1854igmp_qrv - INTEGER
1855	Controls the IGMP query robustness variable (see RFC2236 8.1).
1856
1857	Default: 2 (as specified by RFC2236 8.1)
1858
1859	Minimum: 1 (as specified by RFC6636 4.5)
1860
1861force_igmp_version - INTEGER
1862	- 0 - (default) No enforcement of a IGMP version, IGMPv1/v2 fallback
1863	  allowed. Will back to IGMPv3 mode again if all IGMPv1/v2 Querier
1864	  Present timer expires.
1865	- 1 - Enforce to use IGMP version 1. Will also reply IGMPv1 report if
1866	  receive IGMPv2/v3 query.
1867	- 2 - Enforce to use IGMP version 2. Will fallback to IGMPv1 if receive
1868	  IGMPv1 query message. Will reply report if receive IGMPv3 query.
1869	- 3 - Enforce to use IGMP version 3. The same react with default 0.
1870
1871	.. note::
1872
1873	   this is not the same with force_mld_version because IGMPv3 RFC3376
1874	   Security Considerations does not have clear description that we could
1875	   ignore other version messages completely as MLDv2 RFC3810. So make
1876	   this value as default 0 is recommended.
1877
1878``conf/interface/*``
1879	changes special settings per interface (where
1880	interface" is the name of your network interface)
1881
1882``conf/all/*``
1883	  is special, changes the settings for all interfaces
1884
1885log_martians - BOOLEAN
1886	Log packets with impossible addresses to kernel log.
1887	log_martians for the interface will be enabled if at least one of
1888	conf/{all,interface}/log_martians is set to TRUE,
1889	it will be disabled otherwise
1890
1891accept_redirects - BOOLEAN
1892	Accept ICMP redirect messages.
1893	accept_redirects for the interface will be enabled if:
1894
1895	- both conf/{all,interface}/accept_redirects are TRUE in the case
1896	  forwarding for the interface is enabled
1897
1898	or
1899
1900	- at least one of conf/{all,interface}/accept_redirects is TRUE in the
1901	  case forwarding for the interface is disabled
1902
1903	accept_redirects for the interface will be disabled otherwise
1904
1905	default:
1906
1907		- TRUE (host)
1908		- FALSE (router)
1909
1910forwarding - BOOLEAN
1911	Enable IP forwarding on this interface.  This controls whether packets
1912	received _on_ this interface can be forwarded.
1913
1914mc_forwarding - BOOLEAN
1915	Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
1916	and a multicast routing daemon is required.
1917	conf/all/mc_forwarding must also be set to TRUE to enable multicast
1918	routing	for the interface
1919
1920medium_id - INTEGER
1921	Integer value used to differentiate the devices by the medium they
1922	are attached to. Two devices can have different id values when
1923	the broadcast packets are received only on one of them.
1924	The default value 0 means that the device is the only interface
1925	to its medium, value of -1 means that medium is not known.
1926
1927	Currently, it is used to change the proxy_arp behavior:
1928	the proxy_arp feature is enabled for packets forwarded between
1929	two devices attached to different media.
1930
1931proxy_arp - BOOLEAN
1932	Do proxy arp.
1933
1934	proxy_arp for the interface will be enabled if at least one of
1935	conf/{all,interface}/proxy_arp is set to TRUE,
1936	it will be disabled otherwise
1937
1938proxy_arp_pvlan - BOOLEAN
1939	Private VLAN proxy arp.
1940
1941	Basically allow proxy arp replies back to the same interface
1942	(from which the ARP request/solicitation was received).
1943
1944	This is done to support (ethernet) switch features, like RFC
1945	3069, where the individual ports are NOT allowed to
1946	communicate with each other, but they are allowed to talk to
1947	the upstream router.  As described in RFC 3069, it is possible
1948	to allow these hosts to communicate through the upstream
1949	router by proxy_arp'ing. Don't need to be used together with
1950	proxy_arp.
1951
1952	This technology is known by different names:
1953
1954	- In RFC 3069 it is called VLAN Aggregation.
1955	- Cisco and Allied Telesyn call it Private VLAN.
1956	- Hewlett-Packard call it Source-Port filtering or port-isolation.
1957	- Ericsson call it MAC-Forced Forwarding (RFC Draft).
1958
1959proxy_delay - INTEGER
1960	Delay proxy response.
1961
1962	Delay response to a neighbor solicitation when proxy_arp
1963	or proxy_ndp is enabled. A random value between [0, proxy_delay)
1964	will be chosen, setting to zero means reply with no delay.
1965	Value in jiffies. Defaults to 80.
1966
1967shared_media - BOOLEAN
1968	Send(router) or accept(host) RFC1620 shared media redirects.
1969	Overrides secure_redirects.
1970
1971	shared_media for the interface will be enabled if at least one of
1972	conf/{all,interface}/shared_media is set to TRUE,
1973	it will be disabled otherwise
1974
1975	default TRUE
1976
1977secure_redirects - BOOLEAN
1978	Accept ICMP redirect messages only to gateways listed in the
1979	interface's current gateway list. Even if disabled, RFC1122 redirect
1980	rules still apply.
1981
1982	Overridden by shared_media.
1983
1984	secure_redirects for the interface will be enabled if at least one of
1985	conf/{all,interface}/secure_redirects is set to TRUE,
1986	it will be disabled otherwise
1987
1988	default TRUE
1989
1990send_redirects - BOOLEAN
1991	Send redirects, if router.
1992
1993	send_redirects for the interface will be enabled if at least one of
1994	conf/{all,interface}/send_redirects is set to TRUE,
1995	it will be disabled otherwise
1996
1997	Default: TRUE
1998
1999bootp_relay - BOOLEAN
2000	Accept packets with source address 0.b.c.d destined
2001	not to this host as local ones. It is supposed, that
2002	BOOTP relay daemon will catch and forward such packets.
2003	conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
2004	for the interface
2005
2006	default FALSE
2007
2008	Not Implemented Yet.
2009
2010accept_source_route - BOOLEAN
2011	Accept packets with SRR option.
2012	conf/all/accept_source_route must also be set to TRUE to accept packets
2013	with SRR option on the interface
2014
2015	default
2016
2017		- TRUE (router)
2018		- FALSE (host)
2019
2020accept_local - BOOLEAN
2021	Accept packets with local source addresses. In combination with
2022	suitable routing, this can be used to direct packets between two
2023	local interfaces over the wire and have them accepted properly.
2024	default FALSE
2025
2026route_localnet - BOOLEAN
2027	Do not consider loopback addresses as martian source or destination
2028	while routing. This enables the use of 127/8 for local routing purposes.
2029
2030	default FALSE
2031
2032rp_filter - INTEGER
2033	- 0 - No source validation.
2034	- 1 - Strict mode as defined in RFC3704 Strict Reverse Path
2035	  Each incoming packet is tested against the FIB and if the interface
2036	  is not the best reverse path the packet check will fail.
2037	  By default failed packets are discarded.
2038	- 2 - Loose mode as defined in RFC3704 Loose Reverse Path
2039	  Each incoming packet's source address is also tested against the FIB
2040	  and if the source address is not reachable via any interface
2041	  the packet check will fail.
2042
2043	Current recommended practice in RFC3704 is to enable strict mode
2044	to prevent IP spoofing from DDos attacks. If using asymmetric routing
2045	or other complicated routing, then loose mode is recommended.
2046
2047	The max value from conf/{all,interface}/rp_filter is used
2048	when doing source validation on the {interface}.
2049
2050	Default value is 0. Note that some distributions enable it
2051	in startup scripts.
2052
2053src_valid_mark - BOOLEAN
2054	- 0 - The fwmark of the packet is not included in reverse path
2055	  route lookup.  This allows for asymmetric routing configurations
2056	  utilizing the fwmark in only one direction, e.g., transparent
2057	  proxying.
2058
2059	- 1 - The fwmark of the packet is included in reverse path route
2060	  lookup.  This permits rp_filter to function when the fwmark is
2061	  used for routing traffic in both directions.
2062
2063	This setting also affects the utilization of fmwark when
2064	performing source address selection for ICMP replies, or
2065	determining addresses stored for the IPOPT_TS_TSANDADDR and
2066	IPOPT_RR IP options.
2067
2068	The max value from conf/{all,interface}/src_valid_mark is used.
2069
2070	Default value is 0.
2071
2072arp_filter - BOOLEAN
2073	- 1 - Allows you to have multiple network interfaces on the same
2074	  subnet, and have the ARPs for each interface be answered
2075	  based on whether or not the kernel would route a packet from
2076	  the ARP'd IP out that interface (therefore you must use source
2077	  based routing for this to work). In other words it allows control
2078	  of which cards (usually 1) will respond to an arp request.
2079
2080	- 0 - (default) The kernel can respond to arp requests with addresses
2081	  from other interfaces. This may seem wrong but it usually makes
2082	  sense, because it increases the chance of successful communication.
2083	  IP addresses are owned by the complete host on Linux, not by
2084	  particular interfaces. Only for more complex setups like load-
2085	  balancing, does this behaviour cause problems.
2086
2087	arp_filter for the interface will be enabled if at least one of
2088	conf/{all,interface}/arp_filter is set to TRUE,
2089	it will be disabled otherwise
2090
2091arp_announce - INTEGER
2092	Define different restriction levels for announcing the local
2093	source IP address from IP packets in ARP requests sent on
2094	interface:
2095
2096	- 0 - (default) Use any local address, configured on any interface
2097	- 1 - Try to avoid local addresses that are not in the target's
2098	  subnet for this interface. This mode is useful when target
2099	  hosts reachable via this interface require the source IP
2100	  address in ARP requests to be part of their logical network
2101	  configured on the receiving interface. When we generate the
2102	  request we will check all our subnets that include the
2103	  target IP and will preserve the source address if it is from
2104	  such subnet. If there is no such subnet we select source
2105	  address according to the rules for level 2.
2106	- 2 - Always use the best local address for this target.
2107	  In this mode we ignore the source address in the IP packet
2108	  and try to select local address that we prefer for talks with
2109	  the target host. Such local address is selected by looking
2110	  for primary IP addresses on all our subnets on the outgoing
2111	  interface that include the target IP address. If no suitable
2112	  local address is found we select the first local address
2113	  we have on the outgoing interface or on all other interfaces,
2114	  with the hope we will receive reply for our request and
2115	  even sometimes no matter the source IP address we announce.
2116
2117	The max value from conf/{all,interface}/arp_announce is used.
2118
2119	Increasing the restriction level gives more chance for
2120	receiving answer from the resolved target while decreasing
2121	the level announces more valid sender's information.
2122
2123arp_ignore - INTEGER
2124	Define different modes for sending replies in response to
2125	received ARP requests that resolve local target IP addresses:
2126
2127	- 0 - (default): reply for any local target IP address, configured
2128	  on any interface
2129	- 1 - reply only if the target IP address is local address
2130	  configured on the incoming interface
2131	- 2 - reply only if the target IP address is local address
2132	  configured on the incoming interface and both with the
2133	  sender's IP address are part from same subnet on this interface
2134	- 3 - do not reply for local addresses configured with scope host,
2135	  only resolutions for global and link addresses are replied
2136	- 4-7 - reserved
2137	- 8 - do not reply for all local addresses
2138
2139	The max value from conf/{all,interface}/arp_ignore is used
2140	when ARP request is received on the {interface}
2141
2142arp_notify - BOOLEAN
2143	Define mode for notification of address and device changes.
2144
2145	 ==  ==========================================================
2146	  0  (default): do nothing
2147	  1  Generate gratuitous arp requests when device is brought up
2148	     or hardware address changes.
2149	 ==  ==========================================================
2150
2151arp_accept - INTEGER
2152	Define behavior for accepting gratuitous ARP (garp) frames from devices
2153	that are not already present in the ARP table:
2154
2155	- 0 - don't create new entries in the ARP table
2156	- 1 - create new entries in the ARP table
2157	- 2 - create new entries only if the source IP address is in the same
2158	  subnet as an address configured on the interface that received the
2159	  garp message.
2160
2161	Both replies and requests type gratuitous arp will trigger the
2162	ARP table to be updated, if this setting is on.
2163
2164	If the ARP table already contains the IP address of the
2165	gratuitous arp frame, the arp table will be updated regardless
2166	if this setting is on or off.
2167
2168arp_evict_nocarrier - BOOLEAN
2169	Clears the ARP cache on NOCARRIER events. This option is important for
2170	wireless devices where the ARP cache should not be cleared when roaming
2171	between access points on the same network. In most cases this should
2172	remain as the default (1).
2173
2174	Possible values:
2175
2176	- 0 (disabled) - Do not clear ARP cache on NOCARRIER events
2177	- 1 (enabled)  - Clear the ARP cache on NOCARRIER events
2178
2179	Default: 1 (enabled)
2180
2181mcast_solicit - INTEGER
2182	The maximum number of multicast probes in INCOMPLETE state,
2183	when the associated hardware address is unknown.  Defaults
2184	to 3.
2185
2186ucast_solicit - INTEGER
2187	The maximum number of unicast probes in PROBE state, when
2188	the hardware address is being reconfirmed.  Defaults to 3.
2189
2190app_solicit - INTEGER
2191	The maximum number of probes to send to the user space ARP daemon
2192	via netlink before dropping back to multicast probes (see
2193	mcast_resolicit).  Defaults to 0.
2194
2195mcast_resolicit - INTEGER
2196	The maximum number of multicast probes after unicast and
2197	app probes in PROBE state.  Defaults to 0.
2198
2199disable_policy - BOOLEAN
2200	Disable IPSEC policy (SPD) for this interface
2201
2202	Possible values:
2203
2204	- 0 (disabled)
2205	- 1 (enabled)
2206
2207	Default: 0 (disabled)
2208
2209disable_xfrm - BOOLEAN
2210	Disable IPSEC encryption on this interface, whatever the policy
2211
2212	Possible values:
2213
2214	- 0 (disabled)
2215	- 1 (enabled)
2216
2217	Default: 0 (disabled)
2218
2219igmpv2_unsolicited_report_interval - INTEGER
2220	The interval in milliseconds in which the next unsolicited
2221	IGMPv1 or IGMPv2 report retransmit will take place.
2222
2223	Default: 10000 (10 seconds)
2224
2225igmpv3_unsolicited_report_interval - INTEGER
2226	The interval in milliseconds in which the next unsolicited
2227	IGMPv3 report retransmit will take place.
2228
2229	Default: 1000 (1 seconds)
2230
2231ignore_routes_with_linkdown - BOOLEAN
2232        Ignore routes whose link is down when performing a FIB lookup.
2233
2234        Possible values:
2235
2236	- 0 (disabled)
2237	- 1 (enabled)
2238
2239	Default: 0 (disabled)
2240
2241promote_secondaries - BOOLEAN
2242	When a primary IP address is removed from this interface
2243	promote a corresponding secondary IP address instead of
2244	removing all the corresponding secondary IP addresses.
2245
2246	Possible values:
2247
2248	- 0 (disabled)
2249	- 1 (enabled)
2250
2251	Default: 0 (disabled)
2252
2253drop_unicast_in_l2_multicast - BOOLEAN
2254	Drop any unicast IP packets that are received in link-layer
2255	multicast (or broadcast) frames.
2256
2257	This behavior (for multicast) is actually a SHOULD in RFC
2258	1122, but is disabled by default for compatibility reasons.
2259
2260	Possible values:
2261
2262	- 0 (disabled)
2263	- 1 (enabled)
2264
2265	Default: 0 (disabled)
2266
2267drop_gratuitous_arp - BOOLEAN
2268	Drop all gratuitous ARP frames, for example if there's a known
2269	good ARP proxy on the network and such frames need not be used
2270	(or in the case of 802.11, must not be used to prevent attacks.)
2271
2272	Possible values:
2273
2274	- 0 (disabled)
2275	- 1 (enabled)
2276
2277	Default: 0 (disabled)
2278
2279
2280tag - INTEGER
2281	Allows you to write a number, which can be used as required.
2282
2283	Default value is 0.
2284
2285xfrm4_gc_thresh - INTEGER
2286	(Obsolete since linux-4.14)
2287	The threshold at which we will start garbage collecting for IPv4
2288	destination cache entries.  At twice this value the system will
2289	refuse new allocations.
2290
2291igmp_link_local_mcast_reports - BOOLEAN
2292	Enable IGMP reports for link local multicast groups in the
2293	224.0.0.X range.
2294
2295	Default TRUE
2296
2297Alexey Kuznetsov.
2298kuznet@ms2.inr.ac.ru
2299
2300Updated by:
2301
2302- Andi Kleen
2303  ak@muc.de
2304- Nicolas Delon
2305  delon.nicolas@wanadoo.fr
2306
2307
2308
2309
2310/proc/sys/net/ipv6/* Variables
2311==============================
2312
2313IPv6 has no global variables such as tcp_*.  tcp_* settings under ipv4/ also
2314apply to IPv6 [XXX?].
2315
2316bindv6only - BOOLEAN
2317	Default value for IPV6_V6ONLY socket option,
2318	which restricts use of the IPv6 socket to IPv6 communication
2319	only.
2320
2321	Possible values:
2322
2323	- 0 (disabled) - enable IPv4-mapped address feature
2324	- 1 (enabled)  - disable IPv4-mapped address feature
2325
2326	Default: 0 (disabled)
2327
2328flowlabel_consistency - BOOLEAN
2329	Protect the consistency (and unicity) of flow label.
2330	You have to disable it to use IPV6_FL_F_REFLECT flag on the
2331	flow label manager.
2332
2333	Possible values:
2334
2335	- 0 (disabled)
2336	- 1 (enabled)
2337
2338	Default: 1 (enabled)
2339
2340auto_flowlabels - INTEGER
2341	Automatically generate flow labels based on a flow hash of the
2342	packet. This allows intermediate devices, such as routers, to
2343	identify packet flows for mechanisms like Equal Cost Multipath
2344	Routing (see RFC 6438).
2345
2346	=  ===========================================================
2347	0  automatic flow labels are completely disabled
2348	1  automatic flow labels are enabled by default, they can be
2349	   disabled on a per socket basis using the IPV6_AUTOFLOWLABEL
2350	   socket option
2351	2  automatic flow labels are allowed, they may be enabled on a
2352	   per socket basis using the IPV6_AUTOFLOWLABEL socket option
2353	3  automatic flow labels are enabled and enforced, they cannot
2354	   be disabled by the socket option
2355	=  ===========================================================
2356
2357	Default: 1
2358
2359flowlabel_state_ranges - BOOLEAN
2360	Split the flow label number space into two ranges. 0-0x7FFFF is
2361	reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF
2362	is reserved for stateless flow labels as described in RFC6437.
2363
2364	Possible values:
2365
2366	- 0 (disabled)
2367	- 1 (enabled)
2368
2369	Default: 1 (enabled)
2370
2371
2372flowlabel_reflect - INTEGER
2373	Control flow label reflection. Needed for Path MTU
2374	Discovery to work with Equal Cost Multipath Routing in anycast
2375	environments. See RFC 7690 and:
2376	https://tools.ietf.org/html/draft-wang-6man-flow-label-reflection-01
2377
2378	This is a bitmask.
2379
2380	- 1: enabled for established flows
2381
2382	  Note that this prevents automatic flowlabel changes, as done
2383	  in "tcp: change IPv6 flow-label upon receiving spurious retransmission"
2384	  and "tcp: Change txhash on every SYN and RTO retransmit"
2385
2386	- 2: enabled for TCP RESET packets (no active listener)
2387	  If set, a RST packet sent in response to a SYN packet on a closed
2388	  port will reflect the incoming flow label.
2389
2390	- 4: enabled for ICMPv6 echo reply messages.
2391
2392	Default: 0
2393
2394fib_multipath_hash_policy - INTEGER
2395	Controls which hash policy to use for multipath routes.
2396
2397	Default: 0 (Layer 3)
2398
2399	Possible values:
2400
2401	- 0 - Layer 3 (source and destination addresses plus flow label)
2402	- 1 - Layer 4 (standard 5-tuple)
2403	- 2 - Layer 3 or inner Layer 3 if present
2404	- 3 - Custom multipath hash. Fields used for multipath hash calculation
2405	  are determined by fib_multipath_hash_fields sysctl
2406
2407fib_multipath_hash_fields - UNSIGNED INTEGER
2408	When fib_multipath_hash_policy is set to 3 (custom multipath hash), the
2409	fields used for multipath hash calculation are determined by this
2410	sysctl.
2411
2412	This value is a bitmask which enables various fields for multipath hash
2413	calculation.
2414
2415	Possible fields are:
2416
2417	====== ============================
2418	0x0001 Source IP address
2419	0x0002 Destination IP address
2420	0x0004 IP protocol
2421	0x0008 Flow Label
2422	0x0010 Source port
2423	0x0020 Destination port
2424	0x0040 Inner source IP address
2425	0x0080 Inner destination IP address
2426	0x0100 Inner IP protocol
2427	0x0200 Inner Flow Label
2428	0x0400 Inner source port
2429	0x0800 Inner destination port
2430	====== ============================
2431
2432	Default: 0x0007 (source IP, destination IP and IP protocol)
2433
2434anycast_src_echo_reply - BOOLEAN
2435	Controls the use of anycast addresses as source addresses for ICMPv6
2436	echo reply
2437
2438	Possible values:
2439
2440	- 0 (disabled)
2441	- 1 (enabled)
2442
2443	Default: 0 (disabled)
2444
2445
2446idgen_delay - INTEGER
2447	Controls the delay in seconds after which time to retry
2448	privacy stable address generation if a DAD conflict is
2449	detected.
2450
2451	Default: 1 (as specified in RFC7217)
2452
2453idgen_retries - INTEGER
2454	Controls the number of retries to generate a stable privacy
2455	address if a DAD conflict is detected.
2456
2457	Default: 3 (as specified in RFC7217)
2458
2459mld_qrv - INTEGER
2460	Controls the MLD query robustness variable (see RFC3810 9.1).
2461
2462	Default: 2 (as specified by RFC3810 9.1)
2463
2464	Minimum: 1 (as specified by RFC6636 4.5)
2465
2466max_dst_opts_number - INTEGER
2467	Maximum number of non-padding TLVs allowed in a Destination
2468	options extension header. If this value is less than zero
2469	then unknown options are disallowed and the number of known
2470	TLVs allowed is the absolute value of this number.
2471
2472	Default: 8
2473
2474max_hbh_opts_number - INTEGER
2475	Maximum number of non-padding TLVs allowed in a Hop-by-Hop
2476	options extension header. If this value is less than zero
2477	then unknown options are disallowed and the number of known
2478	TLVs allowed is the absolute value of this number.
2479
2480	Default: 8
2481
2482max_dst_opts_length - INTEGER
2483	Maximum length allowed for a Destination options extension
2484	header.
2485
2486	Default: INT_MAX (unlimited)
2487
2488max_hbh_length - INTEGER
2489	Maximum length allowed for a Hop-by-Hop options extension
2490	header.
2491
2492	Default: INT_MAX (unlimited)
2493
2494skip_notify_on_dev_down - BOOLEAN
2495	Controls whether an RTM_DELROUTE message is generated for routes
2496	removed when a device is taken down or deleted. IPv4 does not
2497	generate this message; IPv6 does by default. Setting this sysctl
2498	to true skips the message, making IPv4 and IPv6 on par in relying
2499	on userspace caches to track link events and evict routes.
2500
2501	Possible values:
2502
2503	- 0 (disabled) - generate the message
2504	- 1 (enabled)  - skip generating the message
2505
2506	Default: 0 (disabled)
2507
2508nexthop_compat_mode - BOOLEAN
2509	New nexthop API provides a means for managing nexthops independent of
2510	prefixes. Backwards compatibility with old route format is enabled by
2511	default which means route dumps and notifications contain the new
2512	nexthop attribute but also the full, expanded nexthop definition.
2513	Further, updates or deletes of a nexthop configuration generate route
2514	notifications for each fib entry using the nexthop. Once a system
2515	understands the new API, this sysctl can be disabled to achieve full
2516	performance benefits of the new API by disabling the nexthop expansion
2517	and extraneous notifications.
2518
2519	Note that as a backward-compatible mode, dumping of modern features
2520	might be incomplete or wrong. For example, resilient groups will not be
2521	shown as such, but rather as just a list of next hops. Also weights that
2522	do not fit into 8 bits will show incorrectly.
2523
2524	Default: true (backward compat mode)
2525
2526fib_notify_on_flag_change - INTEGER
2527        Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/
2528        RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed.
2529
2530        After installing a route to the kernel, user space receives an
2531        acknowledgment, which means the route was installed in the kernel,
2532        but not necessarily in hardware.
2533        It is also possible for a route already installed in hardware to change
2534        its action and therefore its flags. For example, a host route that is
2535        trapping packets can be "promoted" to perform decapsulation following
2536        the installation of an IPinIP/VXLAN tunnel.
2537        The notifications will indicate to user-space the state of the route.
2538
2539        Default: 0 (Do not emit notifications.)
2540
2541        Possible values:
2542
2543        - 0 - Do not emit notifications.
2544        - 1 - Emit notifications.
2545        - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change.
2546
2547ioam6_id - INTEGER
2548        Define the IOAM id of this node. Uses only 24 bits out of 32 in total.
2549
2550        Possible value range:
2551
2552        - Min: 0
2553        - Max: 0xFFFFFF
2554
2555        Default: 0xFFFFFF
2556
2557ioam6_id_wide - LONG INTEGER
2558        Define the wide IOAM id of this node. Uses only 56 bits out of 64 in
2559        total. Can be different from ioam6_id.
2560
2561        Possible value range:
2562
2563        - Min: 0
2564        - Max: 0xFFFFFFFFFFFFFF
2565
2566        Default: 0xFFFFFFFFFFFFFF
2567
2568IPv6 Fragmentation:
2569
2570ip6frag_high_thresh - INTEGER
2571	Maximum memory used to reassemble IPv6 fragments. When
2572	ip6frag_high_thresh bytes of memory is allocated for this purpose,
2573	the fragment handler will toss packets until ip6frag_low_thresh
2574	is reached.
2575
2576ip6frag_low_thresh - INTEGER
2577	See ip6frag_high_thresh
2578
2579ip6frag_time - INTEGER
2580	Time in seconds to keep an IPv6 fragment in memory.
2581
2582``conf/default/*``:
2583	Change the interface-specific default settings.
2584
2585	These settings would be used during creating new interfaces.
2586
2587
2588``conf/all/*``:
2589	Change all the interface-specific settings.
2590
2591	[XXX:  Other special features than forwarding?]
2592
2593conf/all/disable_ipv6 - BOOLEAN
2594	Changing this value is same as changing ``conf/default/disable_ipv6``
2595	setting and also all per-interface ``disable_ipv6`` settings to the same
2596	value.
2597
2598	Reading this value does not have any particular meaning. It does not say
2599	whether IPv6 support is enabled or disabled. Returned value can be 1
2600	also in the case when some interface has ``disable_ipv6`` set to 0 and
2601	has configured IPv6 addresses.
2602
2603conf/all/forwarding - BOOLEAN
2604	Enable global IPv6 forwarding between all interfaces.
2605
2606	IPv4 and IPv6 work differently here; the ``force_forwarding`` flag must
2607	be used to control which interfaces may forward packets.
2608
2609	This also sets all interfaces' Host/Router setting
2610	'forwarding' to the specified value.  See below for details.
2611
2612	This referred to as global forwarding.
2613
2614proxy_ndp - BOOLEAN
2615	Do proxy ndp.
2616
2617	Possible values:
2618
2619	- 0 (disabled)
2620	- 1 (enabled)
2621
2622	Default: 0 (disabled)
2623
2624force_forwarding - BOOLEAN
2625	Enable forwarding on this interface only -- regardless of the setting on
2626	``conf/all/forwarding``. When setting ``conf.all.forwarding`` to 0,
2627	the ``force_forwarding`` flag will be reset on all interfaces.
2628
2629fwmark_reflect - BOOLEAN
2630	Controls the fwmark of kernel-generated IPv6 reply packets that are not
2631	associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
2632	If disabled, these packets have a fwmark of zero. If enabled, they have the
2633	fwmark of the packet they are replying to.
2634
2635	Possible values:
2636
2637	- 0 (disabled)
2638	- 1 (enabled)
2639
2640	Default: 0 (disabled)
2641
2642``conf/interface/*``:
2643	Change special settings per interface.
2644
2645	The functional behaviour for certain settings is different
2646	depending on whether local forwarding is enabled or not.
2647
2648accept_ra - INTEGER
2649	Accept Router Advertisements; autoconfigure using them.
2650
2651	It also determines whether or not to transmit Router
2652	Solicitations. If and only if the functional setting is to
2653	accept Router Advertisements, Router Solicitations will be
2654	transmitted.
2655
2656	Possible values are:
2657
2658		==  ===========================================================
2659		 0  Do not accept Router Advertisements.
2660		 1  Accept Router Advertisements if forwarding is disabled.
2661		 2  Overrule forwarding behaviour. Accept Router Advertisements
2662		    even if forwarding is enabled.
2663		==  ===========================================================
2664
2665	Functional default:
2666
2667		- enabled if local forwarding is disabled.
2668		- disabled if local forwarding is enabled.
2669
2670accept_ra_defrtr - BOOLEAN
2671	Learn default router in Router Advertisement.
2672
2673	Functional default:
2674
2675		- enabled if accept_ra is enabled.
2676		- disabled if accept_ra is disabled.
2677
2678ra_defrtr_metric - UNSIGNED INTEGER
2679	Route metric for default route learned in Router Advertisement. This value
2680	will be assigned as metric for the default route learned via IPv6 Router
2681	Advertisement. Takes affect only if accept_ra_defrtr is enabled.
2682
2683	Possible values:
2684		1 to 0xFFFFFFFF
2685
2686		Default: IP6_RT_PRIO_USER i.e. 1024.
2687
2688accept_ra_from_local - BOOLEAN
2689	Accept RA with source-address that is found on local machine
2690	if the RA is otherwise proper and able to be accepted.
2691
2692	Default is to NOT accept these as it may be an un-intended
2693	network loop.
2694
2695	Functional default:
2696
2697	   - enabled if accept_ra_from_local is enabled
2698	     on a specific interface.
2699	   - disabled if accept_ra_from_local is disabled
2700	     on a specific interface.
2701
2702accept_ra_min_hop_limit - INTEGER
2703	Minimum hop limit Information in Router Advertisement.
2704
2705	Hop limit Information in Router Advertisement less than this
2706	variable shall be ignored.
2707
2708	Default: 1
2709
2710accept_ra_min_lft - INTEGER
2711	Minimum acceptable lifetime value in Router Advertisement.
2712
2713	RA sections with a lifetime less than this value shall be
2714	ignored. Zero lifetimes stay unaffected.
2715
2716	Default: 0
2717
2718accept_ra_pinfo - BOOLEAN
2719	Learn Prefix Information in Router Advertisement.
2720
2721	Functional default:
2722
2723		- enabled if accept_ra is enabled.
2724		- disabled if accept_ra is disabled.
2725
2726ra_honor_pio_life - BOOLEAN
2727	Whether to use RFC4862 Section 5.5.3e to determine the valid
2728	lifetime of an address matching a prefix sent in a Router
2729	Advertisement Prefix Information Option.
2730
2731	Possible values:
2732
2733	- 0 (disabled) - RFC4862 section 5.5.3e is used to determine
2734	  the valid lifetime of the address.
2735	- 1 (enabled)  - the PIO valid lifetime will always be honored.
2736
2737	Default: 0 (disabled)
2738
2739ra_honor_pio_pflag - BOOLEAN
2740	The Prefix Information Option P-flag indicates the network can
2741	allocate a unique IPv6 prefix per client using DHCPv6-PD.
2742	This sysctl can be enabled when a userspace DHCPv6-PD client
2743	is running to cause the P-flag to take effect: i.e. the
2744	P-flag suppresses any effects of the A-flag within the same
2745	PIO. For a given PIO, P=1 and A=1 is treated as A=0.
2746
2747	Possible values:
2748
2749	- 0 (disabled) - the P-flag is ignored.
2750	- 1 (enabled)  - the P-flag will disable SLAAC autoconfiguration
2751	  for the given Prefix Information Option.
2752
2753	Default: 0 (disabled)
2754
2755accept_ra_rt_info_min_plen - INTEGER
2756	Minimum prefix length of Route Information in RA.
2757
2758	Route Information w/ prefix smaller than this variable shall
2759	be ignored.
2760
2761	Functional default:
2762
2763		* 0 if accept_ra_rtr_pref is enabled.
2764		* -1 if accept_ra_rtr_pref is disabled.
2765
2766accept_ra_rt_info_max_plen - INTEGER
2767	Maximum prefix length of Route Information in RA.
2768
2769	Route Information w/ prefix larger than this variable shall
2770	be ignored.
2771
2772	Functional default:
2773
2774		* 0 if accept_ra_rtr_pref is enabled.
2775		* -1 if accept_ra_rtr_pref is disabled.
2776
2777accept_ra_rtr_pref - BOOLEAN
2778	Accept Router Preference in RA.
2779
2780	Functional default:
2781
2782		- enabled if accept_ra is enabled.
2783		- disabled if accept_ra is disabled.
2784
2785accept_ra_mtu - BOOLEAN
2786	Apply the MTU value specified in RA option 5 (RFC4861). If
2787	disabled, the MTU specified in the RA will be ignored.
2788
2789	Functional default:
2790
2791		- enabled if accept_ra is enabled.
2792		- disabled if accept_ra is disabled.
2793
2794accept_redirects - BOOLEAN
2795	Accept Redirects.
2796
2797	Functional default:
2798
2799		- enabled if local forwarding is disabled.
2800		- disabled if local forwarding is enabled.
2801
2802accept_source_route - INTEGER
2803	Accept source routing (routing extension header).
2804
2805	- >= 0: Accept only routing header type 2.
2806	- < 0: Do not accept routing header.
2807
2808	Default: 0
2809
2810autoconf - BOOLEAN
2811	Autoconfigure addresses using Prefix Information in Router
2812	Advertisements.
2813
2814	Functional default:
2815
2816		- enabled if accept_ra_pinfo is enabled.
2817		- disabled if accept_ra_pinfo is disabled.
2818
2819dad_transmits - INTEGER
2820	The amount of Duplicate Address Detection probes to send.
2821
2822	Default: 1
2823
2824forwarding - INTEGER
2825	Configure interface-specific Host/Router behaviour.
2826
2827	.. note::
2828
2829	   It is recommended to have the same setting on all
2830	   interfaces; mixed router/host scenarios are rather uncommon.
2831
2832	Possible values are:
2833
2834		- 0 Forwarding disabled
2835		- 1 Forwarding enabled
2836
2837	**FALSE (0)**:
2838
2839	By default, Host behaviour is assumed.  This means:
2840
2841	1. IsRouter flag is not set in Neighbour Advertisements.
2842	2. If accept_ra is TRUE (default), transmit Router
2843	   Solicitations.
2844	3. If accept_ra is TRUE (default), accept Router
2845	   Advertisements (and do autoconfiguration).
2846	4. If accept_redirects is TRUE (default), accept Redirects.
2847
2848	**TRUE (1)**:
2849
2850	If local forwarding is enabled, Router behaviour is assumed.
2851	This means exactly the reverse from the above:
2852
2853	1. IsRouter flag is set in Neighbour Advertisements.
2854	2. Router Solicitations are not sent unless accept_ra is 2.
2855	3. Router Advertisements are ignored unless accept_ra is 2.
2856	4. Redirects are ignored.
2857
2858	Default: 0 (disabled) if global forwarding is disabled (default),
2859	otherwise 1 (enabled).
2860
2861hop_limit - INTEGER
2862	Default Hop Limit to set.
2863
2864	Default: 64
2865
2866mtu - INTEGER
2867	Default Maximum Transfer Unit
2868
2869	Default: 1280 (IPv6 required minimum)
2870
2871ip_nonlocal_bind - BOOLEAN
2872	If enabled, allows processes to bind() to non-local IPv6 addresses,
2873	which can be quite useful - but may break some applications.
2874
2875	Possible values:
2876
2877	- 0 (disabled)
2878	- 1 (enabled)
2879
2880	Default: 0 (disabled)
2881
2882router_probe_interval - INTEGER
2883	Minimum interval (in seconds) between Router Probing described
2884	in RFC4191.
2885
2886	Default: 60
2887
2888router_solicitation_delay - INTEGER
2889	Number of seconds to wait after interface is brought up
2890	before sending Router Solicitations.
2891
2892	Default: 1
2893
2894router_solicitation_interval - INTEGER
2895	Number of seconds to wait between Router Solicitations.
2896
2897	Default: 4
2898
2899router_solicitations - INTEGER
2900	Number of Router Solicitations to send until assuming no
2901	routers are present.
2902
2903	Default: 3
2904
2905use_oif_addrs_only - BOOLEAN
2906	When enabled, the candidate source addresses for destinations
2907	routed via this interface are restricted to the set of addresses
2908	configured on this interface (vis. RFC 6724, section 4).
2909
2910	Possible values:
2911
2912	- 0 (disabled)
2913	- 1 (enabled)
2914
2915	Default: 0 (disabled)
2916
2917use_tempaddr - INTEGER
2918	Preference for Privacy Extensions (RFC3041).
2919
2920	  * <= 0 : disable Privacy Extensions
2921	  * == 1 : enable Privacy Extensions, but prefer public
2922	    addresses over temporary addresses.
2923	  * >  1 : enable Privacy Extensions and prefer temporary
2924	    addresses over public addresses.
2925
2926	Default:
2927
2928		* 0 (for most devices)
2929		* -1 (for point-to-point devices and loopback devices)
2930
2931temp_valid_lft - INTEGER
2932	valid lifetime (in seconds) for temporary addresses. If less than the
2933	minimum required lifetime (typically 5-7 seconds), temporary addresses
2934	will not be created.
2935
2936	Default: 172800 (2 days)
2937
2938temp_prefered_lft - INTEGER
2939	Preferred lifetime (in seconds) for temporary addresses. If
2940	temp_prefered_lft is less than the minimum required lifetime (typically
2941	5-7 seconds), the preferred lifetime is the minimum required. If
2942	temp_prefered_lft is greater than temp_valid_lft, the preferred lifetime
2943	is temp_valid_lft.
2944
2945	Default: 86400 (1 day)
2946
2947keep_addr_on_down - INTEGER
2948	Keep all IPv6 addresses on an interface down event. If set static
2949	global addresses with no expiration time are not flushed.
2950
2951	*   >0 : enabled
2952	*    0 : system default
2953	*   <0 : disabled
2954
2955	Default: 0 (addresses are removed)
2956
2957max_desync_factor - INTEGER
2958	Maximum value for DESYNC_FACTOR, which is a random value
2959	that ensures that clients don't synchronize with each
2960	other and generate new addresses at exactly the same time.
2961	value is in seconds.
2962
2963	Default: 600
2964
2965regen_min_advance - INTEGER
2966	How far in advance (in seconds), at minimum, to create a new temporary
2967	address before the current one is deprecated. This value is added to
2968	the amount of time that may be required for duplicate address detection
2969	to determine when to create a new address. Linux permits setting this
2970	value to less than the default of 2 seconds, but a value less than 2
2971	does not conform to RFC 8981.
2972
2973	Default: 2
2974
2975regen_max_retry - INTEGER
2976	Number of attempts before give up attempting to generate
2977	valid temporary addresses.
2978
2979	Default: 5
2980
2981max_addresses - INTEGER
2982	Maximum number of autoconfigured addresses per interface.  Setting
2983	to zero disables the limitation.  It is not recommended to set this
2984	value too large (or to zero) because it would be an easy way to
2985	crash the kernel by allowing too many addresses to be created.
2986
2987	Default: 16
2988
2989disable_ipv6 - BOOLEAN
2990	Disable IPv6 operation.  If accept_dad is set to 2, this value
2991	will be dynamically set to TRUE if DAD fails for the link-local
2992	address.
2993
2994	Default: FALSE (enable IPv6 operation)
2995
2996	When this value is changed from 1 to 0 (IPv6 is being enabled),
2997	it will dynamically create a link-local address on the given
2998	interface and start Duplicate Address Detection, if necessary.
2999
3000	When this value is changed from 0 to 1 (IPv6 is being disabled),
3001	it will dynamically delete all addresses and routes on the given
3002	interface. From now on it will not possible to add addresses/routes
3003	to the selected interface.
3004
3005accept_dad - INTEGER
3006	Whether to accept DAD (Duplicate Address Detection).
3007
3008	 == ==============================================================
3009	  0  Disable DAD
3010	  1  Enable DAD (default)
3011	  2  Enable DAD, and disable IPv6 operation if MAC-based duplicate
3012	     link-local address has been found.
3013	 == ==============================================================
3014
3015	DAD operation and mode on a given interface will be selected according
3016	to the maximum value of conf/{all,interface}/accept_dad.
3017
3018force_tllao - BOOLEAN
3019	Enable sending the target link-layer address option even when
3020	responding to a unicast neighbor solicitation.
3021
3022	Default: FALSE
3023
3024	Quoting from RFC 2461, section 4.4, Target link-layer address:
3025
3026	"The option MUST be included for multicast solicitations in order to
3027	avoid infinite Neighbor Solicitation "recursion" when the peer node
3028	does not have a cache entry to return a Neighbor Advertisements
3029	message.  When responding to unicast solicitations, the option can be
3030	omitted since the sender of the solicitation has the correct link-
3031	layer address; otherwise it would not have be able to send the unicast
3032	solicitation in the first place. However, including the link-layer
3033	address in this case adds little overhead and eliminates a potential
3034	race condition where the sender deletes the cached link-layer address
3035	prior to receiving a response to a previous solicitation."
3036
3037ndisc_notify - BOOLEAN
3038	Define mode for notification of address and device changes.
3039
3040	Possible values:
3041
3042	- 0 (disabled) - do nothing
3043	- 1 (enabled)  - Generate unsolicited neighbour advertisements when device is brought
3044	  up or hardware address changes.
3045
3046	Default: 0 (disabled)
3047
3048ndisc_tclass - INTEGER
3049	The IPv6 Traffic Class to use by default when sending IPv6 Neighbor
3050	Discovery (Router Solicitation, Router Advertisement, Neighbor
3051	Solicitation, Neighbor Advertisement, Redirect) messages.
3052	These 8 bits can be interpreted as 6 high order bits holding the DSCP
3053	value and 2 low order bits representing ECN (which you probably want
3054	to leave cleared).
3055
3056	* 0 - (default)
3057
3058ndisc_evict_nocarrier - BOOLEAN
3059	Clears the neighbor discovery table on NOCARRIER events. This option is
3060	important for wireless devices where the neighbor discovery cache should
3061	not be cleared when roaming between access points on the same network.
3062	In most cases this should remain as the default (1).
3063
3064	Possible values:
3065
3066	- 0 (disabled) - Do not clear neighbor discovery cache on NOCARRIER events.
3067	- 1 (enabled)  - Clear neighbor discover cache on NOCARRIER events.
3068
3069	Default: 1 (enabled)
3070
3071mldv1_unsolicited_report_interval - INTEGER
3072	The interval in milliseconds in which the next unsolicited
3073	MLDv1 report retransmit will take place.
3074
3075	Default: 10000 (10 seconds)
3076
3077mldv2_unsolicited_report_interval - INTEGER
3078	The interval in milliseconds in which the next unsolicited
3079	MLDv2 report retransmit will take place.
3080
3081	Default: 1000 (1 second)
3082
3083force_mld_version - INTEGER
3084	* 0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
3085	* 1 - Enforce to use MLD version 1
3086	* 2 - Enforce to use MLD version 2
3087
3088suppress_frag_ndisc - INTEGER
3089	Control RFC 6980 (Security Implications of IPv6 Fragmentation
3090	with IPv6 Neighbor Discovery) behavior:
3091
3092	* 1 - (default) discard fragmented neighbor discovery packets
3093	* 0 - allow fragmented neighbor discovery packets
3094
3095optimistic_dad - BOOLEAN
3096	Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
3097
3098	Optimistic Duplicate Address Detection for the interface will be enabled
3099	if at least one of conf/{all,interface}/optimistic_dad is set to 1,
3100	it will be disabled otherwise.
3101
3102	Possible values:
3103
3104	- 0 (disabled)
3105	- 1 (enabled)
3106
3107	Default: 0 (disabled)
3108
3109
3110use_optimistic - BOOLEAN
3111	If enabled, do not classify optimistic addresses as deprecated during
3112	source address selection.  Preferred addresses will still be chosen
3113	before optimistic addresses, subject to other ranking in the source
3114	address selection algorithm.
3115
3116	This will be enabled if at least one of
3117	conf/{all,interface}/use_optimistic is set to 1, disabled otherwise.
3118
3119	Possible values:
3120
3121	- 0 (disabled)
3122	- 1 (enabled)
3123
3124	Default: 0 (disabled)
3125
3126stable_secret - IPv6 address
3127	This IPv6 address will be used as a secret to generate IPv6
3128	addresses for link-local addresses and autoconfigured
3129	ones. All addresses generated after setting this secret will
3130	be stable privacy ones by default. This can be changed via the
3131	addrgenmode ip-link. conf/default/stable_secret is used as the
3132	secret for the namespace, the interface specific ones can
3133	overwrite that. Writes to conf/all/stable_secret are refused.
3134
3135	It is recommended to generate this secret during installation
3136	of a system and keep it stable after that.
3137
3138	By default the stable secret is unset.
3139
3140addr_gen_mode - INTEGER
3141	Defines how link-local and autoconf addresses are generated.
3142
3143	=  =================================================================
3144	0  generate address based on EUI64 (default)
3145	1  do no generate a link-local address, use EUI64 for addresses
3146	   generated from autoconf
3147	2  generate stable privacy addresses, using the secret from
3148	   stable_secret (RFC7217)
3149	3  generate stable privacy addresses, using a random secret if unset
3150	=  =================================================================
3151
3152drop_unicast_in_l2_multicast - BOOLEAN
3153	Drop any unicast IPv6 packets that are received in link-layer
3154	multicast (or broadcast) frames.
3155
3156	Possible values:
3157
3158	- 0 (disabled)
3159	- 1 (enabled)
3160
3161	Default: 0 (disabled)
3162
3163drop_unsolicited_na - BOOLEAN
3164	Drop all unsolicited neighbor advertisements, for example if there's
3165	a known good NA proxy on the network and such frames need not be used
3166	(or in the case of 802.11, must not be used to prevent attacks.)
3167
3168	Possible values:
3169
3170	- 0 (disabled)
3171	- 1 (enabled)
3172
3173	Default: 0 (disabled).
3174
3175accept_untracked_na - INTEGER
3176	Define behavior for accepting neighbor advertisements from devices that
3177	are absent in the neighbor cache:
3178
3179	- 0 - (default) Do not accept unsolicited and untracked neighbor
3180	  advertisements.
3181
3182	- 1 - Add a new neighbor cache entry in STALE state for routers on
3183	  receiving a neighbor advertisement (either solicited or unsolicited)
3184	  with target link-layer address option specified if no neighbor entry
3185	  is already present for the advertised IPv6 address. Without this knob,
3186	  NAs received for untracked addresses (absent in neighbor cache) are
3187	  silently ignored.
3188
3189	  This is as per router-side behavior documented in RFC9131.
3190
3191	  This has lower precedence than drop_unsolicited_na.
3192
3193	  This will optimize the return path for the initial off-link
3194	  communication that is initiated by a directly connected host, by
3195	  ensuring that the first-hop router which turns on this setting doesn't
3196	  have to buffer the initial return packets to do neighbor-solicitation.
3197	  The prerequisite is that the host is configured to send unsolicited
3198	  neighbor advertisements on interface bringup. This setting should be
3199	  used in conjunction with the ndisc_notify setting on the host to
3200	  satisfy this prerequisite.
3201
3202	- 2 - Extend option (1) to add a new neighbor cache entry only if the
3203	  source IP address is in the same subnet as an address configured on
3204	  the interface that received the neighbor advertisement.
3205
3206enhanced_dad - BOOLEAN
3207	Include a nonce option in the IPv6 neighbor solicitation messages used for
3208	duplicate address detection per RFC7527. A received DAD NS will only signal
3209	a duplicate address if the nonce is different. This avoids any false
3210	detection of duplicates due to loopback of the NS messages that we send.
3211	The nonce option will be sent on an interface unless both of
3212	conf/{all,interface}/enhanced_dad are set to FALSE.
3213
3214	Possible values:
3215
3216	- 0 (disabled)
3217	- 1 (enabled)
3218
3219	Default: 1 (enabled)
3220
3221``icmp/*``:
3222===========
3223
3224ratelimit - INTEGER
3225	Limit the maximal rates for sending ICMPv6 messages.
3226
3227	0 to disable any limiting,
3228	otherwise the minimal space between responses in milliseconds.
3229
3230	Default: 1000
3231
3232ratemask - list of comma separated ranges
3233	For ICMPv6 message types matching the ranges in the ratemask, limit
3234	the sending of the message according to ratelimit parameter.
3235
3236	The format used for both input and output is a comma separated
3237	list of ranges (e.g. "0-127,129" for ICMPv6 message type 0 to 127 and
3238	129). Writing to the file will clear all previous ranges of ICMPv6
3239	message types and update the current list with the input.
3240
3241	Refer to: https://www.iana.org/assignments/icmpv6-parameters/icmpv6-parameters.xhtml
3242	for numerical values of ICMPv6 message types, e.g. echo request is 128
3243	and echo reply is 129.
3244
3245	Default: 0-1,3-127 (rate limit ICMPv6 errors except Packet Too Big)
3246
3247echo_ignore_all - BOOLEAN
3248	If enabled, then the kernel will ignore all ICMP ECHO
3249	requests sent to it over the IPv6 protocol.
3250
3251	Possible values:
3252
3253	- 0 (disabled)
3254	- 1 (enabled)
3255
3256	Default: 0 (disabled)
3257
3258echo_ignore_multicast - BOOLEAN
3259	If enabled, then the kernel will ignore all ICMP ECHO
3260	requests sent to it over the IPv6 protocol via multicast.
3261
3262	Possible values:
3263
3264	- 0 (disabled)
3265	- 1 (enabled)
3266
3267	Default: 0 (disabled)
3268
3269echo_ignore_anycast - BOOLEAN
3270	If enabled, then the kernel will ignore all ICMP ECHO
3271	requests sent to it over the IPv6 protocol destined to anycast address.
3272
3273	Possible values:
3274
3275	- 0 (disabled)
3276	- 1 (enabled)
3277
3278	Default: 0 (disabled)
3279
3280error_anycast_as_unicast - BOOLEAN
3281	If enabled, then the kernel will respond with ICMP Errors
3282	resulting from requests sent to it over the IPv6 protocol destined
3283	to anycast address essentially treating anycast as unicast.
3284
3285	Possible values:
3286
3287	- 0 (disabled)
3288	- 1 (enabled)
3289
3290	Default: 0 (disabled)
3291
3292errors_extension_mask - UNSIGNED INTEGER
3293	Bitmask of ICMP extensions to append to ICMPv6 error messages
3294	("Destination Unreachable" and "Time Exceeded"). The original datagram
3295	is trimmed / padded to 128 bytes in order to be compatible with
3296	applications that do not comply with RFC 4884.
3297
3298	Possible extensions are:
3299
3300	==== ==============================================================
3301	0x01 Incoming IP interface information according to RFC 5837.
3302	     Extension will include the index, IPv6 address (if present),
3303	     name and MTU of the IP interface that received the datagram
3304	     which elicited the ICMP error.
3305	==== ==============================================================
3306
3307	Default: 0x00 (no extensions)
3308
3309xfrm6_gc_thresh - INTEGER
3310	(Obsolete since linux-4.14)
3311	The threshold at which we will start garbage collecting for IPv6
3312	destination cache entries.  At twice this value the system will
3313	refuse new allocations.
3314
3315
3316IPv6 Update by:
3317Pekka Savola <pekkas@netcore.fi>
3318YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
3319
3320
3321/proc/sys/net/bridge/* Variables:
3322=================================
3323
3324bridge-nf-call-arptables - BOOLEAN
3325
3326	Possible values:
3327
3328	- 0 (disabled) - disable this.
3329	- 1 (enabled)  - pass bridged ARP traffic to arptables' FORWARD chain.
3330
3331	Default: 1 (enabled)
3332
3333bridge-nf-call-iptables - BOOLEAN
3334
3335	Possible values:
3336
3337	- 0 (disabled) - disable this.
3338	- 1 (enabled)  - pass bridged IPv4 traffic to iptables' chains.
3339
3340	Default: 1 (enabled)
3341
3342bridge-nf-call-ip6tables - BOOLEAN
3343
3344	Possible values:
3345
3346	- 0 (disabled) - disable this.
3347	- 1 (enabled)  - pass bridged IPv6 traffic to ip6tables' chains.
3348
3349	Default: 1 (enabled)
3350
3351bridge-nf-filter-vlan-tagged - BOOLEAN
3352
3353	Possible values:
3354
3355	- 0 (disabled) - disable this.
3356	- 1 (enabled)  - pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables
3357
3358	Default: 0 (disabled)
3359
3360bridge-nf-filter-pppoe-tagged - BOOLEAN
3361
3362	Possible values:
3363
3364	- 0 (disabled) - disable this.
3365	- 1 (enabled)  - pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
3366
3367	Default: 0 (disabled)
3368
3369bridge-nf-pass-vlan-input-dev - BOOLEAN
3370	- 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
3371	  interface on the bridge and set the netfilter input device to the
3372	  vlan. This allows use of e.g. "iptables -i br0.1" and makes the
3373	  REDIRECT target work with vlan-on-top-of-bridge interfaces.  When no
3374	  matching vlan interface is found, or this switch is off, the input
3375	  device is set to the bridge interface.
3376
3377	- 0: disable bridge netfilter vlan interface lookup.
3378
3379	Default: 0
3380
3381``proc/sys/net/sctp/*`` Variables:
3382==================================
3383
3384addip_enable - BOOLEAN
3385	Enable or disable extension of  Dynamic Address Reconfiguration
3386	(ADD-IP) functionality specified in RFC5061.  This extension provides
3387	the ability to dynamically add and remove new addresses for the SCTP
3388	associations.
3389
3390	Possible values:
3391
3392	- 0 (disabled) - disable extension.
3393	- 1 (enabled)  - enable extension
3394
3395	Default: 0 (disabled)
3396
3397pf_enable - INTEGER
3398	Enable or disable pf (pf is short for potentially failed) state. A value
3399	of pf_retrans > path_max_retrans also disables pf state. That is, one of
3400	both pf_enable and pf_retrans > path_max_retrans can disable pf state.
3401	Since pf_retrans and path_max_retrans can be changed by userspace
3402	application, sometimes user expects to disable pf state by the value of
3403	pf_retrans > path_max_retrans, but occasionally the value of pf_retrans
3404	or path_max_retrans is changed by the user application, this pf state is
3405	enabled. As such, it is necessary to add this to dynamically enable
3406	and disable pf state. See:
3407	https://datatracker.ietf.org/doc/draft-ietf-tsvwg-sctp-failover for
3408	details.
3409
3410	Possible values:
3411
3412	- 1: Enable pf.
3413	- 0: Disable pf.
3414
3415	Default: 1
3416
3417pf_expose - INTEGER
3418	Unset or enable/disable pf (pf is short for potentially failed) state
3419	exposure.  Applications can control the exposure of the PF path state
3420	in the SCTP_PEER_ADDR_CHANGE event and access of SCTP_PF-state
3421	transport info via SCTP_GET_PEER_ADDR_INFO sockopt.
3422
3423	Possible values:
3424
3425	- 0: Unset pf state exposure (compatible with old applications). No
3426	  event will be sent but the transport info can be queried.
3427	- 1: Disable pf state exposure. No event will be sent and trying to
3428	  obtain transport info will return -EACCESS.
3429	- 2: Enable pf state exposure. The event will be sent for a transport
3430	  becoming SCTP_PF state and transport info can be obtained.
3431
3432	Default: 0
3433
3434addip_noauth_enable - BOOLEAN
3435	Dynamic Address Reconfiguration (ADD-IP) requires the use of
3436	authentication to protect the operations of adding or removing new
3437	addresses.  This requirement is mandated so that unauthorized hosts
3438	would not be able to hijack associations.  However, older
3439	implementations may not have implemented this requirement while
3440	allowing the ADD-IP extension.  For reasons of interoperability,
3441	we provide this variable to control the enforcement of the
3442	authentication requirement.
3443
3444	== ===============================================================
3445	1  Allow ADD-IP extension to be used without authentication.  This
3446	   should only be set in a closed environment for interoperability
3447	   with older implementations.
3448
3449	0  Enforce the authentication requirement
3450	== ===============================================================
3451
3452	Default: 0
3453
3454auth_enable - BOOLEAN
3455	Enable or disable Authenticated Chunks extension.  This extension
3456	provides the ability to send and receive authenticated chunks and is
3457	required for secure operation of Dynamic Address Reconfiguration
3458	(ADD-IP) extension.
3459
3460	Possible values:
3461
3462	- 0 (disabled) - disable extension.
3463	- 1 (enabled)  - enable extension
3464
3465	Default: 0 (disabled)
3466
3467prsctp_enable - BOOLEAN
3468	Enable or disable the Partial Reliability extension (RFC3758) which
3469	is used to notify peers that a given DATA should no longer be expected.
3470
3471	Possible values:
3472
3473	- 0 (disabled) - disable extension.
3474	- 1 (enabled)  - enable extension
3475
3476	Default: 1 (enabled)
3477
3478max_burst - INTEGER
3479	The limit of the number of new packets that can be initially sent.  It
3480	controls how bursty the generated traffic can be.
3481
3482	Default: 4
3483
3484association_max_retrans - INTEGER
3485	Set the maximum number for retransmissions that an association can
3486	attempt deciding that the remote end is unreachable.  If this value
3487	is exceeded, the association is terminated.
3488
3489	Default: 10
3490
3491max_init_retransmits - INTEGER
3492	The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
3493	that an association will attempt before declaring the destination
3494	unreachable and terminating.
3495
3496	Default: 8
3497
3498path_max_retrans - INTEGER
3499	The maximum number of retransmissions that will be attempted on a given
3500	path.  Once this threshold is exceeded, the path is considered
3501	unreachable, and new traffic will use a different path when the
3502	association is multihomed.
3503
3504	Default: 5
3505
3506pf_retrans - INTEGER
3507	The number of retransmissions that will be attempted on a given path
3508	before traffic is redirected to an alternate transport (should one
3509	exist).  Note this is distinct from path_max_retrans, as a path that
3510	passes the pf_retrans threshold can still be used.  Its only
3511	deprioritized when a transmission path is selected by the stack.  This
3512	setting is primarily used to enable fast failover mechanisms without
3513	having to reduce path_max_retrans to a very low value.  See:
3514	http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
3515	for details.  Note also that a value of pf_retrans > path_max_retrans
3516	disables this feature. Since both pf_retrans and path_max_retrans can
3517	be changed by userspace application, a variable pf_enable is used to
3518	disable pf state.
3519
3520	Default: 0
3521
3522ps_retrans - INTEGER
3523	Primary.Switchover.Max.Retrans (PSMR), it's a tunable parameter coming
3524	from section-5 "Primary Path Switchover" in rfc7829.  The primary path
3525	will be changed to another active path when the path error counter on
3526	the old primary path exceeds PSMR, so that "the SCTP sender is allowed
3527	to continue data transmission on a new working path even when the old
3528	primary destination address becomes active again".   Note this feature
3529	is disabled by initializing 'ps_retrans' per netns as 0xffff by default,
3530	and its value can't be less than 'pf_retrans' when changing by sysctl.
3531
3532	Default: 0xffff
3533
3534rto_initial - INTEGER
3535	The initial round trip timeout value in milliseconds that will be used
3536	in calculating round trip times.  This is the initial time interval
3537	for retransmissions.
3538
3539	Default: 3000
3540
3541rto_max - INTEGER
3542	The maximum value (in milliseconds) of the round trip timeout.  This
3543	is the largest time interval that can elapse between retransmissions.
3544
3545	Default: 60000
3546
3547rto_min - INTEGER
3548	The minimum value (in milliseconds) of the round trip timeout.  This
3549	is the smallest time interval the can elapse between retransmissions.
3550
3551	Default: 1000
3552
3553hb_interval - INTEGER
3554	The interval (in milliseconds) between HEARTBEAT chunks.  These chunks
3555	are sent at the specified interval on idle paths to probe the state of
3556	a given path between 2 associations.
3557
3558	Default: 30000
3559
3560sack_timeout - INTEGER
3561	The amount of time (in milliseconds) that the implementation will wait
3562	to send a SACK.
3563
3564	Default: 200
3565
3566valid_cookie_life - INTEGER
3567	The default lifetime of the SCTP cookie (in milliseconds).  The cookie
3568	is used during association establishment.
3569
3570	Default: 60000
3571
3572cookie_preserve_enable - BOOLEAN
3573	Enable or disable the ability to extend the lifetime of the SCTP cookie
3574	that is used during the establishment phase of SCTP association
3575
3576	Possible values:
3577
3578	- 0 (disabled) - disable.
3579	- 1 (enabled)  - enable cookie lifetime extension.
3580
3581	Default: 1 (enabled)
3582
3583cookie_hmac_alg - STRING
3584	Select the hmac algorithm used when generating the cookie value sent by
3585	a listening sctp socket to a connecting client in the INIT-ACK chunk.
3586	Valid values are:
3587
3588	* sha256
3589	* none
3590
3591	Default: sha256
3592
3593rcvbuf_policy - INTEGER
3594	Determines if the receive buffer is attributed to the socket or to
3595	association.   SCTP supports the capability to create multiple
3596	associations on a single socket.  When using this capability, it is
3597	possible that a single stalled association that's buffering a lot
3598	of data may block other associations from delivering their data by
3599	consuming all of the receive buffer space.  To work around this,
3600	the rcvbuf_policy could be set to attribute the receiver buffer space
3601	to each association instead of the socket.  This prevents the described
3602	blocking.
3603
3604	- 1: rcvbuf space is per association
3605	- 0: rcvbuf space is per socket
3606
3607	Default: 0
3608
3609sndbuf_policy - INTEGER
3610	Similar to rcvbuf_policy above, this applies to send buffer space.
3611
3612	- 1: Send buffer is tracked per association
3613	- 0: Send buffer is tracked per socket.
3614
3615	Default: 0
3616
3617sctp_mem - vector of 3 INTEGERs: min, pressure, max
3618	Number of pages allowed for queueing by all SCTP sockets.
3619
3620	* min: Below this number of pages SCTP is not bothered about its
3621	  memory usage. When amount of memory allocated by SCTP exceeds
3622	  this number, SCTP starts to moderate memory usage.
3623	* pressure: This value was introduced to follow format of tcp_mem.
3624	* max: Maximum number of allowed pages.
3625
3626	Default is calculated at boot time from amount of available memory.
3627
3628sctp_rmem - vector of 3 INTEGERs: min, default, max
3629	Only the first value ("min") is used, "default" and "max" are
3630	ignored.
3631
3632	* min: Minimal size of receive buffer used by SCTP socket.
3633	  It is guaranteed to each SCTP socket (but not association) even
3634	  under moderate memory pressure.
3635
3636	Default: 4K
3637
3638sctp_wmem  - vector of 3 INTEGERs: min, default, max
3639	Only the first value ("min") is used, "default" and "max" are
3640	ignored.
3641
3642	* min: Minimum size of send buffer that can be used by SCTP sockets.
3643	  It is guaranteed to each SCTP socket (but not association) even
3644	  under moderate memory pressure.
3645
3646	Default: 4K
3647
3648addr_scope_policy - INTEGER
3649	Control IPv4 address scoping (see
3650	https://datatracker.ietf.org/doc/draft-stewart-tsvwg-sctp-ipv4/00/
3651	for details).
3652
3653	- 0   - Disable IPv4 address scoping
3654	- 1   - Enable IPv4 address scoping
3655	- 2   - Follow draft but allow IPv4 private addresses
3656	- 3   - Follow draft but allow IPv4 link local addresses
3657
3658	Default: 1
3659
3660udp_port - INTEGER
3661	The listening port for the local UDP tunneling sock. Normally it's
3662	using the IANA-assigned UDP port number 9899 (sctp-tunneling).
3663
3664	This UDP sock is used for processing the incoming UDP-encapsulated
3665	SCTP packets (from RFC6951), and shared by all applications in the
3666	same net namespace. This UDP sock will be closed when the value is
3667	set to 0.
3668
3669	The value will also be used to set the src port of the UDP header
3670	for the outgoing UDP-encapsulated SCTP packets. For the dest port,
3671	please refer to 'encap_port' below.
3672
3673	Default: 0
3674
3675encap_port - INTEGER
3676	The default remote UDP encapsulation port.
3677
3678	This value is used to set the dest port of the UDP header for the
3679	outgoing UDP-encapsulated SCTP packets by default. Users can also
3680	change the value for each sock/asoc/transport by using setsockopt.
3681	For further information, please refer to RFC6951.
3682
3683	Note that when connecting to a remote server, the client should set
3684	this to the port that the UDP tunneling sock on the peer server is
3685	listening to and the local UDP tunneling sock on the client also
3686	must be started. On the server, it would get the encap_port from
3687	the incoming packet's source port.
3688
3689	Default: 0
3690
3691plpmtud_probe_interval - INTEGER
3692        The time interval (in milliseconds) for the PLPMTUD probe timer,
3693        which is configured to expire after this period to receive an
3694        acknowledgment to a probe packet. This is also the time interval
3695        between the probes for the current pmtu when the probe search
3696        is done.
3697
3698        PLPMTUD will be disabled when 0 is set, and other values for it
3699        must be >= 5000.
3700
3701	Default: 0
3702
3703reconf_enable - BOOLEAN
3704        Enable or disable extension of Stream Reconfiguration functionality
3705        specified in RFC6525. This extension provides the ability to "reset"
3706        a stream, and it includes the Parameters of "Outgoing/Incoming SSN
3707        Reset", "SSN/TSN Reset" and "Add Outgoing/Incoming Streams".
3708
3709	Possible values:
3710
3711	- 0 (disabled) - Disable extension.
3712	- 1 (enabled) - Enable extension.
3713
3714	Default: 0 (disabled)
3715
3716intl_enable - BOOLEAN
3717        Enable or disable extension of User Message Interleaving functionality
3718        specified in RFC8260. This extension allows the interleaving of user
3719        messages sent on different streams. With this feature enabled, I-DATA
3720        chunk will replace DATA chunk to carry user messages if also supported
3721        by the peer. Note that to use this feature, one needs to set this option
3722        to 1 and also needs to set socket options SCTP_FRAGMENT_INTERLEAVE to 2
3723        and SCTP_INTERLEAVING_SUPPORTED to 1.
3724
3725	Possible values:
3726
3727	- 0 (disabled) - Disable extension.
3728	- 1 (enabled) - Enable extension.
3729
3730	Default: 0 (disabled)
3731
3732ecn_enable - BOOLEAN
3733        Control use of Explicit Congestion Notification (ECN) by SCTP.
3734        Like in TCP, ECN is used only when both ends of the SCTP connection
3735        indicate support for it. This feature is useful in avoiding losses
3736        due to congestion by allowing supporting routers to signal congestion
3737        before having to drop packets.
3738
3739        Possible values:
3740
3741	- 0 (disabled) - Disable ecn.
3742	- 1 (enabled) - Enable ecn.
3743
3744	Default: 1 (enabled)
3745
3746l3mdev_accept - BOOLEAN
3747	Enabling this option allows a "global" bound socket to work
3748	across L3 master domains (e.g., VRFs) with packets capable of
3749	being received regardless of the L3 domain in which they
3750	originated. Only valid when the kernel was compiled with
3751	CONFIG_NET_L3_MASTER_DEV.
3752
3753	Possible values:
3754
3755	- 0 (disabled)
3756	- 1 (enabled)
3757
3758	Default: 1 (enabled)
3759
3760
3761``/proc/sys/net/core/*``
3762========================
3763
3764	Please see: Documentation/admin-guide/sysctl/net.rst for descriptions of these entries.
3765
3766
3767``/proc/sys/net/unix/*``
3768========================
3769
3770max_dgram_qlen - INTEGER
3771	The maximum length of dgram socket receive queue
3772
3773	Default: 10
3774
3775