1/proc/sys/net/ipv4/* Variables:
   2
   3ip_forward - BOOLEAN
   4	0 - disabled (default)
   5	not 0 - enabled
   6
   7	Forward Packets between interfaces.
   8
   9	This variable is special, its change resets all configuration
  10	parameters to their default state (RFC1122 for hosts, RFC1812
  11	for routers)
  12
  13ip_default_ttl - INTEGER
  14	Default value of TTL field (Time To Live) for outgoing (but not
  15	forwarded) IP packets. Should be between 1 and 255 inclusive.
  16	Default: 64 (as recommended by RFC1700)
  17
  18ip_no_pmtu_disc - INTEGER
  19	Disable Path MTU Discovery. If enabled in mode 1 and a
  20	fragmentation-required ICMP is received, the PMTU to this
  21	destination will be set to min_pmtu (see below). You will need
  22	to raise min_pmtu to the smallest interface MTU on your system
  23	manually if you want to avoid locally generated fragments.
  24
  25	In mode 2 incoming Path MTU Discovery messages will be
  26	discarded. Outgoing frames are handled the same as in mode 1,
  27	implicitly setting IP_PMTUDISC_DONT on every created socket.
  28
  29	Mode 3 is a hardend pmtu discover mode. The kernel will only
  30	accept fragmentation-needed errors if the underlying protocol
  31	can verify them besides a plain socket lookup. Current
  32	protocols for which pmtu events will be honored are TCP, SCTP
  33	and DCCP as they verify e.g. the sequence number or the
  34	association. This mode should not be enabled globally but is
  35	only intended to secure e.g. name servers in namespaces where
  36	TCP path mtu must still work but path MTU information of other
  37	protocols should be discarded. If enabled globally this mode
  38	could break other protocols.
  39
  40	Possible values: 0-3
  41	Default: FALSE
  42
  43min_pmtu - INTEGER
  44	default 552 - minimum discovered Path MTU
  45
  46ip_forward_use_pmtu - BOOLEAN
  47	By default we don't trust protocol path MTUs while forwarding
  48	because they could be easily forged and can lead to unwanted
  49	fragmentation by the router.
  50	You only need to enable this if you have user-space software
  51	which tries to discover path mtus by itself and depends on the
  52	kernel honoring this information. This is normally not the
  53	case.
  54	Default: 0 (disabled)
  55	Possible values:
  56	0 - disabled
  57	1 - enabled
  58
  59fwmark_reflect - BOOLEAN
  60	Controls the fwmark of kernel-generated IPv4 reply packets that are not
  61	associated with a socket for example, TCP RSTs or ICMP echo replies).
  62	If unset, these packets have a fwmark of zero. If set, they have the
  63	fwmark of the packet they are replying to.
  64	Default: 0
  65
  66fib_multipath_use_neigh - BOOLEAN
  67	Use status of existing neighbor entry when determining nexthop for
  68	multipath routes. If disabled, neighbor information is not used and
  69	packets could be directed to a failed nexthop. Only valid for kernels
  70	built with CONFIG_IP_ROUTE_MULTIPATH enabled.
  71	Default: 0 (disabled)
  72	Possible values:
  73	0 - disabled
  74	1 - enabled
  75
  76route/max_size - INTEGER
  77	Maximum number of routes allowed in the kernel.  Increase
  78	this when using large numbers of interfaces and/or routes.
  79	From linux kernel 3.6 onwards, this is deprecated for ipv4
  80	as route cache is no longer used.
  81
  82neigh/default/gc_thresh1 - INTEGER
  83	Minimum number of entries to keep.  Garbage collector will not
  84	purge entries if there are fewer than this number.
  85	Default: 128
  86
  87neigh/default/gc_thresh2 - INTEGER
  88	Threshold when garbage collector becomes more aggressive about
  89	purging entries. Entries older than 5 seconds will be cleared
  90	when over this number.
  91	Default: 512
  92
  93neigh/default/gc_thresh3 - INTEGER
  94	Maximum number of neighbor entries allowed.  Increase this
  95	when using large numbers of interfaces and when communicating
  96	with large numbers of directly-connected peers.
  97	Default: 1024
  98
  99neigh/default/unres_qlen_bytes - INTEGER
 100	The maximum number of bytes which may be used by packets
 101	queued for each	unresolved address by other network layers.
 102	(added in linux 3.3)
 103	Setting negative value is meaningless and will return error.
 104	Default: 65536 Bytes(64KB)
 105
 106neigh/default/unres_qlen - INTEGER
 107	The maximum number of packets which may be queued for each
 108	unresolved address by other network layers.
 109	(deprecated in linux 3.3) : use unres_qlen_bytes instead.
 110	Prior to linux 3.3, the default value is 3 which may cause
 111	unexpected packet loss. The current default value is calculated
 112	according to default value of unres_qlen_bytes and true size of
 113	packet.
 114	Default: 31
 115
 116mtu_expires - INTEGER
 117	Time, in seconds, that cached PMTU information is kept.
 118
 119min_adv_mss - INTEGER
 120	The advertised MSS depends on the first hop route MTU, but will
 121	never be lower than this setting.
 122
 123IP Fragmentation:
 124
 125ipfrag_high_thresh - INTEGER
 126	Maximum memory used to reassemble IP fragments. When
 127	ipfrag_high_thresh bytes of memory is allocated for this purpose,
 128	the fragment handler will toss packets until ipfrag_low_thresh
 129	is reached. This also serves as a maximum limit to namespaces
 130	different from the initial one.
 131
 132ipfrag_low_thresh - INTEGER
 133	Maximum memory used to reassemble IP fragments before the kernel
 134	begins to remove incomplete fragment queues to free up resources.
 135	The kernel still accepts new fragments for defragmentation.
 136
 137ipfrag_time - INTEGER
 138	Time in seconds to keep an IP fragment in memory.
 139
 140ipfrag_max_dist - INTEGER
 141	ipfrag_max_dist is a non-negative integer value which defines the
 142	maximum "disorder" which is allowed among fragments which share a
 143	common IP source address. Note that reordering of packets is
 144	not unusual, but if a large number of fragments arrive from a source
 145	IP address while a particular fragment queue remains incomplete, it
 146	probably indicates that one or more fragments belonging to that queue
 147	have been lost. When ipfrag_max_dist is positive, an additional check
 148	is done on fragments before they are added to a reassembly queue - if
 149	ipfrag_max_dist (or more) fragments have arrived from a particular IP
 150	address between additions to any IP fragment queue using that source
 151	address, it's presumed that one or more fragments in the queue are
 152	lost. The existing fragment queue will be dropped, and a new one
 153	started. An ipfrag_max_dist value of zero disables this check.
 154
 155	Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
 156	result in unnecessarily dropping fragment queues when normal
 157	reordering of packets occurs, which could lead to poor application
 158	performance. Using a very large value, e.g. 50000, increases the
 159	likelihood of incorrectly reassembling IP fragments that originate
 160	from different IP datagrams, which could result in data corruption.
 161	Default: 64
 162
 163INET peer storage:
 164
 165inet_peer_threshold - INTEGER
 166	The approximate size of the storage.  Starting from this threshold
 167	entries will be thrown aggressively.  This threshold also determines
 168	entries' time-to-live and time intervals between garbage collection
 169	passes.  More entries, less time-to-live, less GC interval.
 170
 171inet_peer_minttl - INTEGER
 172	Minimum time-to-live of entries.  Should be enough to cover fragment
 173	time-to-live on the reassembling side.  This minimum time-to-live  is
 174	guaranteed if the pool size is less than inet_peer_threshold.
 175	Measured in seconds.
 176
 177inet_peer_maxttl - INTEGER
 178	Maximum time-to-live of entries.  Unused entries will expire after
 179	this period of time if there is no memory pressure on the pool (i.e.
 180	when the number of entries in the pool is very small).
 181	Measured in seconds.
 182
 183TCP variables:
 184
 185somaxconn - INTEGER
 186	Limit of socket listen() backlog, known in userspace as SOMAXCONN.
 187	Defaults to 128.  See also tcp_max_syn_backlog for additional tuning
 188	for TCP sockets.
 189
 190tcp_abort_on_overflow - BOOLEAN
 191	If listening service is too slow to accept new connections,
 192	reset them. Default state is FALSE. It means that if overflow
 193	occurred due to a burst, connection will recover. Enable this
 194	option _only_ if you are really sure that listening daemon
 195	cannot be tuned to accept connections faster. Enabling this
 196	option can harm clients of your server.
 197
 198tcp_adv_win_scale - INTEGER
 199	Count buffering overhead as bytes/2^tcp_adv_win_scale
 200	(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
 201	if it is <= 0.
 202	Possible values are [-31, 31], inclusive.
 203	Default: 1
 204
 205tcp_allowed_congestion_control - STRING
 206	Show/set the congestion control choices available to non-privileged
 207	processes. The list is a subset of those listed in
 208	tcp_available_congestion_control.
 209	Default is "reno" and the default setting (tcp_congestion_control).
 210
 211tcp_app_win - INTEGER
 212	Reserve max(window/2^tcp_app_win, mss) of window for application
 213	buffer. Value 0 is special, it means that nothing is reserved.
 214	Default: 31
 215
 216tcp_autocorking - BOOLEAN
 217	Enable TCP auto corking :
 218	When applications do consecutive small write()/sendmsg() system calls,
 219	we try to coalesce these small writes as much as possible, to lower
 220	total amount of sent packets. This is done if at least one prior
 221	packet for the flow is waiting in Qdisc queues or device transmit
 222	queue. Applications can still use TCP_CORK for optimal behavior
 223	when they know how/when to uncork their sockets.
 224	Default : 1
 225
 226tcp_available_congestion_control - STRING
 227	Shows the available congestion control choices that are registered.
 228	More congestion control algorithms may be available as modules,
 229	but not loaded.
 230
 231tcp_base_mss - INTEGER
 232	The initial value of search_low to be used by the packetization layer
 233	Path MTU discovery (MTU probing).  If MTU probing is enabled,
 234	this is the initial MSS used by the connection.
 235
 236tcp_congestion_control - STRING
 237	Set the congestion control algorithm to be used for new
 238	connections. The algorithm "reno" is always available, but
 239	additional choices may be available based on kernel configuration.
 240	Default is set as part of kernel configuration.
 241	For passive connections, the listener congestion control choice
 242	is inherited.
 243	[see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
 244
 245tcp_dsack - BOOLEAN
 246	Allows TCP to send "duplicate" SACKs.
 247
 248tcp_early_retrans - INTEGER
 249	Enable Early Retransmit (ER), per RFC 5827. ER lowers the threshold
 250	for triggering fast retransmit when the amount of outstanding data is
 251	small and when no previously unsent data can be transmitted (such
 252	that limited transmit could be used). Also controls the use of
 253	Tail loss probe (TLP) that converts RTOs occurring due to tail
 254	losses into fast recovery (draft-dukkipati-tcpm-tcp-loss-probe-01).
 255	Possible values:
 256		0 disables ER
 257		1 enables ER
 258		2 enables ER but delays fast recovery and fast retransmit
 259		  by a fourth of RTT. This mitigates connection falsely
 260		  recovers when network has a small degree of reordering
 261		  (less than 3 packets).
 262		3 enables delayed ER and TLP.
 263		4 enables TLP only.
 264	Default: 3
 265
 266tcp_ecn - INTEGER
 267	Control use of Explicit Congestion Notification (ECN) by TCP.
 268	ECN is used only when both ends of the TCP connection indicate
 269	support for it.  This feature is useful in avoiding losses due
 270	to congestion by allowing supporting routers to signal
 271	congestion before having to drop packets.
 272	Possible values are:
 273		0 Disable ECN.  Neither initiate nor accept ECN.
 274		1 Enable ECN when requested by incoming connections and
 275		  also request ECN on outgoing connection attempts.
 276		2 Enable ECN when requested by incoming connections
 277		  but do not request ECN on outgoing connections.
 278	Default: 2
 279
 280tcp_ecn_fallback - BOOLEAN
 281	If the kernel detects that ECN connection misbehaves, enable fall
 282	back to non-ECN. Currently, this knob implements the fallback
 283	from RFC3168, section 6.1.1.1., but we reserve that in future,
 284	additional detection mechanisms could be implemented under this
 285	knob. The value	is not used, if tcp_ecn or per route (or congestion
 286	control) ECN settings are disabled.
 287	Default: 1 (fallback enabled)
 288
 289tcp_fack - BOOLEAN
 290	Enable FACK congestion avoidance and fast retransmission.
 291	The value is not used, if tcp_sack is not enabled.
 292
 293tcp_fin_timeout - INTEGER
 294	The length of time an orphaned (no longer referenced by any
 295	application) connection will remain in the FIN_WAIT_2 state
 296	before it is aborted at the local end.  While a perfectly
 297	valid "receive only" state for an un-orphaned connection, an
 298	orphaned connection in FIN_WAIT_2 state could otherwise wait
 299	forever for the remote to close its end of the connection.
 300	Cf. tcp_max_orphans
 301	Default: 60 seconds
 302
 303tcp_frto - INTEGER
 304	Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
 305	F-RTO is an enhanced recovery algorithm for TCP retransmission
 306	timeouts.  It is particularly beneficial in networks where the
 307	RTT fluctuates (e.g., wireless). F-RTO is sender-side only
 308	modification. It does not require any support from the peer.
 309
 310	By default it's enabled with a non-zero value. 0 disables F-RTO.
 311
 312tcp_invalid_ratelimit - INTEGER
 313	Limit the maximal rate for sending duplicate acknowledgments
 314	in response to incoming TCP packets that are for an existing
 315	connection but that are invalid due to any of these reasons:
 316
 317	  (a) out-of-window sequence number,
 318	  (b) out-of-window acknowledgment number, or
 319	  (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
 320
 321	This can help mitigate simple "ack loop" DoS attacks, wherein
 322	a buggy or malicious middlebox or man-in-the-middle can
 323	rewrite TCP header fields in manner that causes each endpoint
 324	to think that the other is sending invalid TCP segments, thus
 325	causing each side to send an unterminating stream of duplicate
 326	acknowledgments for invalid segments.
 327
 328	Using 0 disables rate-limiting of dupacks in response to
 329	invalid segments; otherwise this value specifies the minimal
 330	space between sending such dupacks, in milliseconds.
 331
 332	Default: 500 (milliseconds).
 333
 334tcp_keepalive_time - INTEGER
 335	How often TCP sends out keepalive messages when keepalive is enabled.
 336	Default: 2hours.
 337
 338tcp_keepalive_probes - INTEGER
 339	How many keepalive probes TCP sends out, until it decides that the
 340	connection is broken. Default value: 9.
 341
 342tcp_keepalive_intvl - INTEGER
 343	How frequently the probes are send out. Multiplied by
 344	tcp_keepalive_probes it is time to kill not responding connection,
 345	after probes started. Default value: 75sec i.e. connection
 346	will be aborted after ~11 minutes of retries.
 347
 348tcp_l3mdev_accept - BOOLEAN
 349	Enables child sockets to inherit the L3 master device index.
 350	Enabling this option allows a "global" listen socket to work
 351	across L3 master domains (e.g., VRFs) with connected sockets
 352	derived from the listen socket to be bound to the L3 domain in
 353	which the packets originated. Only valid when the kernel was
 354	compiled with CONFIG_NET_L3_MASTER_DEV.
 355
 356tcp_low_latency - BOOLEAN
 357	If set, the TCP stack makes decisions that prefer lower
 358	latency as opposed to higher throughput.  By default, this
 359	option is not set meaning that higher throughput is preferred.
 360	An example of an application where this default should be
 361	changed would be a Beowulf compute cluster.
 362	Default: 0
 363
 364tcp_max_orphans - INTEGER
 365	Maximal number of TCP sockets not attached to any user file handle,
 366	held by system.	If this number is exceeded orphaned connections are
 367	reset immediately and warning is printed. This limit exists
 368	only to prevent simple DoS attacks, you _must_ not rely on this
 369	or lower the limit artificially, but rather increase it
 370	(probably, after increasing installed memory),
 371	if network conditions require more than default value,
 372	and tune network services to linger and kill such states
 373	more aggressively. Let me to remind again: each orphan eats
 374	up to ~64K of unswappable memory.
 375
 376tcp_max_syn_backlog - INTEGER
 377	Maximal number of remembered connection requests, which have not
 378	received an acknowledgment from connecting client.
 379	The minimal value is 128 for low memory machines, and it will
 380	increase in proportion to the memory of machine.
 381	If server suffers from overload, try increasing this number.
 382
 383tcp_max_tw_buckets - INTEGER
 384	Maximal number of timewait sockets held by system simultaneously.
 385	If this number is exceeded time-wait socket is immediately destroyed
 386	and warning is printed. This limit exists only to prevent
 387	simple DoS attacks, you _must_ not lower the limit artificially,
 388	but rather increase it (probably, after increasing installed memory),
 389	if network conditions require more than default value.
 390
 391tcp_mem - vector of 3 INTEGERs: min, pressure, max
 392	min: below this number of pages TCP is not bothered about its
 393	memory appetite.
 394
 395	pressure: when amount of memory allocated by TCP exceeds this number
 396	of pages, TCP moderates its memory consumption and enters memory
 397	pressure mode, which is exited when memory consumption falls
 398	under "min".
 399
 400	max: number of pages allowed for queueing by all TCP sockets.
 401
 402	Defaults are calculated at boot time from amount of available
 403	memory.
 404
 405tcp_min_rtt_wlen - INTEGER
 406	The window length of the windowed min filter to track the minimum RTT.
 407	A shorter window lets a flow more quickly pick up new (higher)
 408	minimum RTT when it is moved to a longer path (e.g., due to traffic
 409	engineering). A longer window makes the filter more resistant to RTT
 410	inflations such as transient congestion. The unit is seconds.
 411	Default: 300
 412
 413tcp_moderate_rcvbuf - BOOLEAN
 414	If set, TCP performs receive buffer auto-tuning, attempting to
 415	automatically size the buffer (no greater than tcp_rmem[2]) to
 416	match the size required by the path for full throughput.  Enabled by
 417	default.
 418
 419tcp_mtu_probing - INTEGER
 420	Controls TCP Packetization-Layer Path MTU Discovery.  Takes three
 421	values:
 422	  0 - Disabled
 423	  1 - Disabled by default, enabled when an ICMP black hole detected
 424	  2 - Always enabled, use initial MSS of tcp_base_mss.
 425
 426tcp_probe_interval - INTEGER
 427	Controls how often to start TCP Packetization-Layer Path MTU
 428	Discovery reprobe. The default is reprobing every 10 minutes as
 429	per RFC4821.
 430
 431tcp_probe_threshold - INTEGER
 432	Controls when TCP Packetization-Layer Path MTU Discovery probing
 433	will stop in respect to the width of search range in bytes. Default
 434	is 8 bytes.
 435
 436tcp_no_metrics_save - BOOLEAN
 437	By default, TCP saves various connection metrics in the route cache
 438	when the connection closes, so that connections established in the
 439	near future can use these to set initial conditions.  Usually, this
 440	increases overall performance, but may sometimes cause performance
 441	degradation.  If set, TCP will not cache metrics on closing
 442	connections.
 443
 444tcp_orphan_retries - INTEGER
 445	This value influences the timeout of a locally closed TCP connection,
 446	when RTO retransmissions remain unacknowledged.
 447	See tcp_retries2 for more details.
 448
 449	The default value is 8.
 450	If your machine is a loaded WEB server,
 451	you should think about lowering this value, such sockets
 452	may consume significant resources. Cf. tcp_max_orphans.
 453
 454tcp_recovery - INTEGER
 455	This value is a bitmap to enable various experimental loss recovery
 456	features.
 457
 458	RACK: 0x1 enables the RACK loss detection for fast detection of lost
 459	      retransmissions and tail drops.
 460
 461	Default: 0x1
 462
 463tcp_reordering - INTEGER
 464	Initial reordering level of packets in a TCP stream.
 465	TCP stack can then dynamically adjust flow reordering level
 466	between this initial value and tcp_max_reordering
 467	Default: 3
 468
 469tcp_max_reordering - INTEGER
 470	Maximal reordering level of packets in a TCP stream.
 471	300 is a fairly conservative value, but you might increase it
 472	if paths are using per packet load balancing (like bonding rr mode)
 473	Default: 300
 474
 475tcp_retrans_collapse - BOOLEAN
 476	Bug-to-bug compatibility with some broken printers.
 477	On retransmit try to send bigger packets to work around bugs in
 478	certain TCP stacks.
 479
 480tcp_retries1 - INTEGER
 481	This value influences the time, after which TCP decides, that
 482	something is wrong due to unacknowledged RTO retransmissions,
 483	and reports this suspicion to the network layer.
 484	See tcp_retries2 for more details.
 485
 486	RFC 1122 recommends at least 3 retransmissions, which is the
 487	default.
 488
 489tcp_retries2 - INTEGER
 490	This value influences the timeout of an alive TCP connection,
 491	when RTO retransmissions remain unacknowledged.
 492	Given a value of N, a hypothetical TCP connection following
 493	exponential backoff with an initial RTO of TCP_RTO_MIN would
 494	retransmit N times before killing the connection at the (N+1)th RTO.
 495
 496	The default value of 15 yields a hypothetical timeout of 924.6
 497	seconds and is a lower bound for the effective timeout.
 498	TCP will effectively time out at the first RTO which exceeds the
 499	hypothetical timeout.
 500
 501	RFC 1122 recommends at least 100 seconds for the timeout,
 502	which corresponds to a value of at least 8.
 503
 504tcp_rfc1337 - BOOLEAN
 505	If set, the TCP stack behaves conforming to RFC1337. If unset,
 506	we are not conforming to RFC, but prevent TCP TIME_WAIT
 507	assassination.
 508	Default: 0
 509
 510tcp_rmem - vector of 3 INTEGERs: min, default, max
 511	min: Minimal size of receive buffer used by TCP sockets.
 512	It is guaranteed to each TCP socket, even under moderate memory
 513	pressure.
 514	Default: 1 page
 515
 516	default: initial size of receive buffer used by TCP sockets.
 517	This value overrides net.core.rmem_default used by other protocols.
 518	Default: 87380 bytes. This value results in window of 65535 with
 519	default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
 520	less for default tcp_app_win. See below about these variables.
 521
 522	max: maximal size of receive buffer allowed for automatically
 523	selected receiver buffers for TCP socket. This value does not override
 524	net.core.rmem_max.  Calling setsockopt() with SO_RCVBUF disables
 525	automatic tuning of that socket's receive buffer size, in which
 526	case this value is ignored.
 527	Default: between 87380B and 6MB, depending on RAM size.
 528
 529tcp_sack - BOOLEAN
 530	Enable select acknowledgments (SACKS).
 531
 532tcp_slow_start_after_idle - BOOLEAN
 533	If set, provide RFC2861 behavior and time out the congestion
 534	window after an idle period.  An idle period is defined at
 535	the current RTO.  If unset, the congestion window will not
 536	be timed out after an idle period.
 537	Default: 1
 538
 539tcp_stdurg - BOOLEAN
 540	Use the Host requirements interpretation of the TCP urgent pointer field.
 541	Most hosts use the older BSD interpretation, so if you turn this on
 542	Linux might not communicate correctly with them.
 543	Default: FALSE
 544
 545tcp_synack_retries - INTEGER
 546	Number of times SYNACKs for a passive TCP connection attempt will
 547	be retransmitted. Should not be higher than 255. Default value
 548	is 5, which corresponds to 31seconds till the last retransmission
 549	with the current initial RTO of 1second. With this the final timeout
 550	for a passive TCP connection will happen after 63seconds.
 551
 552tcp_syncookies - BOOLEAN
 553	Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
 554	Send out syncookies when the syn backlog queue of a socket
 555	overflows. This is to prevent against the common 'SYN flood attack'
 556	Default: 1
 557
 558	Note, that syncookies is fallback facility.
 559	It MUST NOT be used to help highly loaded servers to stand
 560	against legal connection rate. If you see SYN flood warnings
 561	in your logs, but investigation	shows that they occur
 562	because of overload with legal connections, you should tune
 563	another parameters until this warning disappear.
 564	See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
 565
 566	syncookies seriously violate TCP protocol, do not allow
 567	to use TCP extensions, can result in serious degradation
 568	of some services (f.e. SMTP relaying), visible not by you,
 569	but your clients and relays, contacting you. While you see
 570	SYN flood warnings in logs not being really flooded, your server
 571	is seriously misconfigured.
 572
 573	If you want to test which effects syncookies have to your
 574	network connections you can set this knob to 2 to enable
 575	unconditionally generation of syncookies.
 576
 577tcp_fastopen - INTEGER
 578	Enable TCP Fast Open (RFC7413) to send and accept data in the opening
 579	SYN packet.
 580
 581	The client support is enabled by flag 0x1 (on by default). The client
 582	then must use sendmsg() or sendto() with the MSG_FASTOPEN flag,
 583	rather than connect() to send data in SYN.
 584
 585	The server support is enabled by flag 0x2 (off by default). Then
 586	either enable for all listeners with another flag (0x400) or
 587	enable individual listeners via TCP_FASTOPEN socket option with
 588	the option value being the length of the syn-data backlog.
 589
 590	The values (bitmap) are
 591	  0x1: (client) enables sending data in the opening SYN on the client.
 592	  0x2: (server) enables the server support, i.e., allowing data in
 593			a SYN packet to be accepted and passed to the
 594			application before 3-way handshake finishes.
 595	  0x4: (client) send data in the opening SYN regardless of cookie
 596			availability and without a cookie option.
 597	0x200: (server) accept data-in-SYN w/o any cookie option present.
 598	0x400: (server) enable all listeners to support Fast Open by
 599			default without explicit TCP_FASTOPEN socket option.
 600
 601	Default: 0x1
 602
 603	Note that that additional client or server features are only
 604	effective if the basic support (0x1 and 0x2) are enabled respectively.
 605
 606tcp_syn_retries - INTEGER
 607	Number of times initial SYNs for an active TCP connection attempt
 608	will be retransmitted. Should not be higher than 127. Default value
 609	is 6, which corresponds to 63seconds till the last retransmission
 610	with the current initial RTO of 1second. With this the final timeout
 611	for an active TCP connection attempt will happen after 127seconds.
 612
 613tcp_timestamps - INTEGER
 614Enable timestamps as defined in RFC1323.
 615	0: Disabled.
 616	1: Enable timestamps as defined in RFC1323 and use random offset for
 617	each connection rather than only using the current time.
 618	2: Like 1, but without random offsets.
 619	Default: 1
 620
 621tcp_min_tso_segs - INTEGER
 622	Minimal number of segments per TSO frame.
 623	Since linux-3.12, TCP does an automatic sizing of TSO frames,
 624	depending on flow rate, instead of filling 64Kbytes packets.
 625	For specific usages, it's possible to force TCP to build big
 626	TSO frames. Note that TCP stack might split too big TSO packets
 627	if available window is too small.
 628	Default: 2
 629
 630tcp_pacing_ss_ratio - INTEGER
 631	sk->sk_pacing_rate is set by TCP stack using a ratio applied
 632	to current rate. (current_rate = cwnd * mss / srtt)
 633	If TCP is in slow start, tcp_pacing_ss_ratio is applied
 634	to let TCP probe for bigger speeds, assuming cwnd can be
 635	doubled every other RTT.
 636	Default: 200
 637
 638tcp_pacing_ca_ratio - INTEGER
 639	sk->sk_pacing_rate is set by TCP stack using a ratio applied
 640	to current rate. (current_rate = cwnd * mss / srtt)
 641	If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio
 642	is applied to conservatively probe for bigger throughput.
 643	Default: 120
 644
 645tcp_tso_win_divisor - INTEGER
 646	This allows control over what percentage of the congestion window
 647	can be consumed by a single TSO frame.
 648	The setting of this parameter is a choice between burstiness and
 649	building larger TSO frames.
 650	Default: 3
 651
 652tcp_tw_recycle - BOOLEAN
 653	Enable fast recycling TIME-WAIT sockets. Default value is 0.
 654	It should not be changed without advice/request of technical
 655	experts.
 656
 657tcp_tw_reuse - BOOLEAN
 658	Allow to reuse TIME-WAIT sockets for new connections when it is
 659	safe from protocol viewpoint. Default value is 0.
 660	It should not be changed without advice/request of technical
 661	experts.
 662
 663tcp_window_scaling - BOOLEAN
 664	Enable window scaling as defined in RFC1323.
 665
 666tcp_wmem - vector of 3 INTEGERs: min, default, max
 667	min: Amount of memory reserved for send buffers for TCP sockets.
 668	Each TCP socket has rights to use it due to fact of its birth.
 669	Default: 1 page
 670
 671	default: initial size of send buffer used by TCP sockets.  This
 672	value overrides net.core.wmem_default used by other protocols.
 673	It is usually lower than net.core.wmem_default.
 674	Default: 16K
 675
 676	max: Maximal amount of memory allowed for automatically tuned
 677	send buffers for TCP sockets. This value does not override
 678	net.core.wmem_max.  Calling setsockopt() with SO_SNDBUF disables
 679	automatic tuning of that socket's send buffer size, in which case
 680	this value is ignored.
 681	Default: between 64K and 4MB, depending on RAM size.
 682
 683tcp_notsent_lowat - UNSIGNED INTEGER
 684	A TCP socket can control the amount of unsent bytes in its write queue,
 685	thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
 686	reports POLLOUT events if the amount of unsent bytes is below a per
 687	socket value, and if the write queue is not full. sendmsg() will
 688	also not add new buffers if the limit is hit.
 689
 690	This global variable controls the amount of unsent data for
 691	sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
 692	to the global variable has immediate effect.
 693
 694	Default: UINT_MAX (0xFFFFFFFF)
 695
 696tcp_workaround_signed_windows - BOOLEAN
 697	If set, assume no receipt of a window scaling option means the
 698	remote TCP is broken and treats the window as a signed quantity.
 699	If unset, assume the remote TCP is not broken even if we do
 700	not receive a window scaling option from them.
 701	Default: 0
 702
 703tcp_thin_linear_timeouts - BOOLEAN
 704	Enable dynamic triggering of linear timeouts for thin streams.
 705	If set, a check is performed upon retransmission by timeout to
 706	determine if the stream is thin (less than 4 packets in flight).
 707	As long as the stream is found to be thin, up to 6 linear
 708	timeouts may be performed before exponential backoff mode is
 709	initiated. This improves retransmission latency for
 710	non-aggressive thin streams, often found to be time-dependent.
 711	For more information on thin streams, see
 712	Documentation/networking/tcp-thin.txt
 713	Default: 0
 714
 715tcp_thin_dupack - BOOLEAN
 716	Enable dynamic triggering of retransmissions after one dupACK
 717	for thin streams. If set, a check is performed upon reception
 718	of a dupACK to determine if the stream is thin (less than 4
 719	packets in flight). As long as the stream is found to be thin,
 720	data is retransmitted on the first received dupACK. This
 721	improves retransmission latency for non-aggressive thin
 722	streams, often found to be time-dependent.
 723	For more information on thin streams, see
 724	Documentation/networking/tcp-thin.txt
 725	Default: 0
 726
 727tcp_limit_output_bytes - INTEGER
 728	Controls TCP Small Queue limit per tcp socket.
 729	TCP bulk sender tends to increase packets in flight until it
 730	gets losses notifications. With SNDBUF autotuning, this can
 731	result in a large amount of packets queued in qdisc/device
 732	on the local machine, hurting latency of other flows, for
 733	typical pfifo_fast qdiscs.
 734	tcp_limit_output_bytes limits the number of bytes on qdisc
 735	or device to reduce artificial RTT/cwnd and reduce bufferbloat.
 736	Default: 262144
 737
 738tcp_challenge_ack_limit - INTEGER
 739	Limits number of Challenge ACK sent per second, as recommended
 740	in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
 741	Default: 100
 742
 743UDP variables:
 744
 745udp_mem - vector of 3 INTEGERs: min, pressure, max
 746	Number of pages allowed for queueing by all UDP sockets.
 747
 748	min: Below this number of pages UDP is not bothered about its
 749	memory appetite. When amount of memory allocated by UDP exceeds
 750	this number, UDP starts to moderate memory usage.
 751
 752	pressure: This value was introduced to follow format of tcp_mem.
 753
 754	max: Number of pages allowed for queueing by all UDP sockets.
 755
 756	Default is calculated at boot time from amount of available memory.
 757
 758udp_rmem_min - INTEGER
 759	Minimal size of receive buffer used by UDP sockets in moderation.
 760	Each UDP socket is able to use the size for receiving data, even if
 761	total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
 762	Default: 1 page
 763
 764udp_wmem_min - INTEGER
 765	Minimal size of send buffer used by UDP sockets in moderation.
 766	Each UDP socket is able to use the size for sending data, even if
 767	total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
 768	Default: 1 page
 769
 770CIPSOv4 Variables:
 771
 772cipso_cache_enable - BOOLEAN
 773	If set, enable additions to and lookups from the CIPSO label mapping
 774	cache.  If unset, additions are ignored and lookups always result in a
 775	miss.  However, regardless of the setting the cache is still
 776	invalidated when required when means you can safely toggle this on and
 777	off and the cache will always be "safe".
 778	Default: 1
 779
 780cipso_cache_bucket_size - INTEGER
 781	The CIPSO label cache consists of a fixed size hash table with each
 782	hash bucket containing a number of cache entries.  This variable limits
 783	the number of entries in each hash bucket; the larger the value the
 784	more CIPSO label mappings that can be cached.  When the number of
 785	entries in a given hash bucket reaches this limit adding new entries
 786	causes the oldest entry in the bucket to be removed to make room.
 787	Default: 10
 788
 789cipso_rbm_optfmt - BOOLEAN
 790	Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
 791	the CIPSO draft specification (see Documentation/netlabel for details).
 792	This means that when set the CIPSO tag will be padded with empty
 793	categories in order to make the packet data 32-bit aligned.
 794	Default: 0
 795
 796cipso_rbm_structvalid - BOOLEAN
 797	If set, do a very strict check of the CIPSO option when
 798	ip_options_compile() is called.  If unset, relax the checks done during
 799	ip_options_compile().  Either way is "safe" as errors are caught else
 800	where in the CIPSO processing code but setting this to 0 (False) should
 801	result in less work (i.e. it should be faster) but could cause problems
 802	with other implementations that require strict checking.
 803	Default: 0
 804
 805IP Variables:
 806
 807ip_local_port_range - 2 INTEGERS
 808	Defines the local port range that is used by TCP and UDP to
 809	choose the local port. The first number is the first, the
 810	second the last local port number.
 811	If possible, it is better these numbers have different parity.
 812	(one even and one odd values)
 813	The default values are 32768 and 60999 respectively.
 814
 815ip_local_reserved_ports - list of comma separated ranges
 816	Specify the ports which are reserved for known third-party
 817	applications. These ports will not be used by automatic port
 818	assignments (e.g. when calling connect() or bind() with port
 819	number 0). Explicit port allocation behavior is unchanged.
 820
 821	The format used for both input and output is a comma separated
 822	list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
 823	10). Writing to the file will clear all previously reserved
 824	ports and update the current list with the one given in the
 825	input.
 826
 827	Note that ip_local_port_range and ip_local_reserved_ports
 828	settings are independent and both are considered by the kernel
 829	when determining which ports are available for automatic port
 830	assignments.
 831
 832	You can reserve ports which are not in the current
 833	ip_local_port_range, e.g.:
 834
 835	$ cat /proc/sys/net/ipv4/ip_local_port_range
 836	32000	60999
 837	$ cat /proc/sys/net/ipv4/ip_local_reserved_ports
 838	8080,9148
 839
 840	although this is redundant. However such a setting is useful
 841	if later the port range is changed to a value that will
 842	include the reserved ports.
 843
 844	Default: Empty
 845
 846ip_nonlocal_bind - BOOLEAN
 847	If set, allows processes to bind() to non-local IP addresses,
 848	which can be quite useful - but may break some applications.
 849	Default: 0
 850
 851ip_dynaddr - BOOLEAN
 852	If set non-zero, enables support for dynamic addresses.
 853	If set to a non-zero value larger than 1, a kernel log
 854	message will be printed when dynamic address rewriting
 855	occurs.
 856	Default: 0
 857
 858ip_early_demux - BOOLEAN
 859	Optimize input packet processing down to one demux for
 860	certain kinds of local sockets.  Currently we only do this
 861	for established TCP sockets.
 862
 863	It may add an additional cost for pure routing workloads that
 864	reduces overall throughput, in such case you should disable it.
 865	Default: 1
 866
 867icmp_echo_ignore_all - BOOLEAN
 868	If set non-zero, then the kernel will ignore all ICMP ECHO
 869	requests sent to it.
 870	Default: 0
 871
 872icmp_echo_ignore_broadcasts - BOOLEAN
 873	If set non-zero, then the kernel will ignore all ICMP ECHO and
 874	TIMESTAMP requests sent to it via broadcast/multicast.
 875	Default: 1
 876
 877icmp_ratelimit - INTEGER
 878	Limit the maximal rates for sending ICMP packets whose type matches
 879	icmp_ratemask (see below) to specific targets.
 880	0 to disable any limiting,
 881	otherwise the minimal space between responses in milliseconds.
 882	Note that another sysctl, icmp_msgs_per_sec limits the number
 883	of ICMP packets	sent on all targets.
 884	Default: 1000
 885
 886icmp_msgs_per_sec - INTEGER
 887	Limit maximal number of ICMP packets sent per second from this host.
 888	Only messages whose type matches icmp_ratemask (see below) are
 889	controlled by this limit.
 890	Default: 1000
 891
 892icmp_msgs_burst - INTEGER
 893	icmp_msgs_per_sec controls number of ICMP packets sent per second,
 894	while icmp_msgs_burst controls the burst size of these packets.
 895	Default: 50
 896
 897icmp_ratemask - INTEGER
 898	Mask made of ICMP types for which rates are being limited.
 899	Significant bits: IHGFEDCBA9876543210
 900	Default mask:     0000001100000011000 (6168)
 901
 902	Bit definitions (see include/linux/icmp.h):
 903		0 Echo Reply
 904		3 Destination Unreachable *
 905		4 Source Quench *
 906		5 Redirect
 907		8 Echo Request
 908		B Time Exceeded *
 909		C Parameter Problem *
 910		D Timestamp Request
 911		E Timestamp Reply
 912		F Info Request
 913		G Info Reply
 914		H Address Mask Request
 915		I Address Mask Reply
 916
 917	* These are rate limited by default (see default mask above)
 918
 919icmp_ignore_bogus_error_responses - BOOLEAN
 920	Some routers violate RFC1122 by sending bogus responses to broadcast
 921	frames.  Such violations are normally logged via a kernel warning.
 922	If this is set to TRUE, the kernel will not give such warnings, which
 923	will avoid log file clutter.
 924	Default: 1
 925
 926icmp_errors_use_inbound_ifaddr - BOOLEAN
 927
 928	If zero, icmp error messages are sent with the primary address of
 929	the exiting interface.
 930
 931	If non-zero, the message will be sent with the primary address of
 932	the interface that received the packet that caused the icmp error.
 933	This is the behaviour network many administrators will expect from
 934	a router. And it can make debugging complicated network layouts
 935	much easier.
 936
 937	Note that if no primary address exists for the interface selected,
 938	then the primary address of the first non-loopback interface that
 939	has one will be used regardless of this setting.
 940
 941	Default: 0
 942
 943igmp_max_memberships - INTEGER
 944	Change the maximum number of multicast groups we can subscribe to.
 945	Default: 20
 946
 947	Theoretical maximum value is bounded by having to send a membership
 948	report in a single datagram (i.e. the report can't span multiple
 949	datagrams, or risk confusing the switch and leaving groups you don't
 950	intend to).
 951
 952	The number of supported groups 'M' is bounded by the number of group
 953	report entries you can fit into a single datagram of 65535 bytes.
 954
 955	M = 65536-sizeof (ip header)/(sizeof(Group record))
 956
 957	Group records are variable length, with a minimum of 12 bytes.
 958	So net.ipv4.igmp_max_memberships should not be set higher than:
 959
 960	(65536-24) / 12 = 5459
 961
 962	The value 5459 assumes no IP header options, so in practice
 963	this number may be lower.
 964
 965igmp_max_msf - INTEGER
 966	Maximum number of addresses allowed in the source filter list for a
 967	multicast group.
 968	Default: 10
 969
 970igmp_qrv - INTEGER
 971	Controls the IGMP query robustness variable (see RFC2236 8.1).
 972	Default: 2 (as specified by RFC2236 8.1)
 973	Minimum: 1 (as specified by RFC6636 4.5)
 974
 975force_igmp_version - INTEGER
 976	0 - (default) No enforcement of a IGMP version, IGMPv1/v2 fallback
 977	    allowed. Will back to IGMPv3 mode again if all IGMPv1/v2 Querier
 978	    Present timer expires.
 979	1 - Enforce to use IGMP version 1. Will also reply IGMPv1 report if
 980	    receive IGMPv2/v3 query.
 981	2 - Enforce to use IGMP version 2. Will fallback to IGMPv1 if receive
 982	    IGMPv1 query message. Will reply report if receive IGMPv3 query.
 983	3 - Enforce to use IGMP version 3. The same react with default 0.
 984
 985	Note: this is not the same with force_mld_version because IGMPv3 RFC3376
 986	Security Considerations does not have clear description that we could
 987	ignore other version messages completely as MLDv2 RFC3810. So make
 988	this value as default 0 is recommended.
 989
 990conf/interface/*  changes special settings per interface (where
 991"interface" is the name of your network interface)
 992
 993conf/all/*	  is special, changes the settings for all interfaces
 994
 995log_martians - BOOLEAN
 996	Log packets with impossible addresses to kernel log.
 997	log_martians for the interface will be enabled if at least one of
 998	conf/{all,interface}/log_martians is set to TRUE,
 999	it will be disabled otherwise
1000
1001accept_redirects - BOOLEAN
1002	Accept ICMP redirect messages.
1003	accept_redirects for the interface will be enabled if:
1004	- both conf/{all,interface}/accept_redirects are TRUE in the case
1005	  forwarding for the interface is enabled
1006	or
1007	- at least one of conf/{all,interface}/accept_redirects is TRUE in the
1008	  case forwarding for the interface is disabled
1009	accept_redirects for the interface will be disabled otherwise
1010	default TRUE (host)
1011		FALSE (router)
1012
1013forwarding - BOOLEAN
1014	Enable IP forwarding on this interface.
1015
1016mc_forwarding - BOOLEAN
1017	Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
1018	and a multicast routing daemon is required.
1019	conf/all/mc_forwarding must also be set to TRUE to enable multicast
1020	routing	for the interface
1021
1022medium_id - INTEGER
1023	Integer value used to differentiate the devices by the medium they
1024	are attached to. Two devices can have different id values when
1025	the broadcast packets are received only on one of them.
1026	The default value 0 means that the device is the only interface
1027	to its medium, value of -1 means that medium is not known.
1028
1029	Currently, it is used to change the proxy_arp behavior:
1030	the proxy_arp feature is enabled for packets forwarded between
1031	two devices attached to different media.
1032
1033proxy_arp - BOOLEAN
1034	Do proxy arp.
1035	proxy_arp for the interface will be enabled if at least one of
1036	conf/{all,interface}/proxy_arp is set to TRUE,
1037	it will be disabled otherwise
1038
1039proxy_arp_pvlan - BOOLEAN
1040	Private VLAN proxy arp.
1041	Basically allow proxy arp replies back to the same interface
1042	(from which the ARP request/solicitation was received).
1043
1044	This is done to support (ethernet) switch features, like RFC
1045	3069, where the individual ports are NOT allowed to
1046	communicate with each other, but they are allowed to talk to
1047	the upstream router.  As described in RFC 3069, it is possible
1048	to allow these hosts to communicate through the upstream
1049	router by proxy_arp'ing. Don't need to be used together with
1050	proxy_arp.
1051
1052	This technology is known by different names:
1053	  In RFC 3069 it is called VLAN Aggregation.
1054	  Cisco and Allied Telesyn call it Private VLAN.
1055	  Hewlett-Packard call it Source-Port filtering or port-isolation.
1056	  Ericsson call it MAC-Forced Forwarding (RFC Draft).
1057
1058shared_media - BOOLEAN
1059	Send(router) or accept(host) RFC1620 shared media redirects.
1060	Overrides secure_redirects.
1061	shared_media for the interface will be enabled if at least one of
1062	conf/{all,interface}/shared_media is set to TRUE,
1063	it will be disabled otherwise
1064	default TRUE
1065
1066secure_redirects - BOOLEAN
1067	Accept ICMP redirect messages only to gateways listed in the
1068	interface's current gateway list. Even if disabled, RFC1122 redirect
1069	rules still apply.
1070	Overridden by shared_media.
1071	secure_redirects for the interface will be enabled if at least one of
1072	conf/{all,interface}/secure_redirects is set to TRUE,
1073	it will be disabled otherwise
1074	default TRUE
1075
1076send_redirects - BOOLEAN
1077	Send redirects, if router.
1078	send_redirects for the interface will be enabled if at least one of
1079	conf/{all,interface}/send_redirects is set to TRUE,
1080	it will be disabled otherwise
1081	Default: TRUE
1082
1083bootp_relay - BOOLEAN
1084	Accept packets with source address 0.b.c.d destined
1085	not to this host as local ones. It is supposed, that
1086	BOOTP relay daemon will catch and forward such packets.
1087	conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
1088	for the interface
1089	default FALSE
1090	Not Implemented Yet.
1091
1092accept_source_route - BOOLEAN
1093	Accept packets with SRR option.
1094	conf/all/accept_source_route must also be set to TRUE to accept packets
1095	with SRR option on the interface
1096	default TRUE (router)
1097		FALSE (host)
1098
1099accept_local - BOOLEAN
1100	Accept packets with local source addresses. In combination with
1101	suitable routing, this can be used to direct packets between two
1102	local interfaces over the wire and have them accepted properly.
1103	default FALSE
1104
1105route_localnet - BOOLEAN
1106	Do not consider loopback addresses as martian source or destination
1107	while routing. This enables the use of 127/8 for local routing purposes.
1108	default FALSE
1109
1110rp_filter - INTEGER
1111	0 - No source validation.
1112	1 - Strict mode as defined in RFC3704 Strict Reverse Path
1113	    Each incoming packet is tested against the FIB and if the interface
1114	    is not the best reverse path the packet check will fail.
1115	    By default failed packets are discarded.
1116	2 - Loose mode as defined in RFC3704 Loose Reverse Path
1117	    Each incoming packet's source address is also tested against the FIB
1118	    and if the source address is not reachable via any interface
1119	    the packet check will fail.
1120
1121	Current recommended practice in RFC3704 is to enable strict mode
1122	to prevent IP spoofing from DDos attacks. If using asymmetric routing
1123	or other complicated routing, then loose mode is recommended.
1124
1125	The max value from conf/{all,interface}/rp_filter is used
1126	when doing source validation on the {interface}.
1127
1128	Default value is 0. Note that some distributions enable it
1129	in startup scripts.
1130
1131arp_filter - BOOLEAN
1132	1 - Allows you to have multiple network interfaces on the same
1133	subnet, and have the ARPs for each interface be answered
1134	based on whether or not the kernel would route a packet from
1135	the ARP'd IP out that interface (therefore you must use source
1136	based routing for this to work). In other words it allows control
1137	of which cards (usually 1) will respond to an arp request.
1138
1139	0 - (default) The kernel can respond to arp requests with addresses
1140	from other interfaces. This may seem wrong but it usually makes
1141	sense, because it increases the chance of successful communication.
1142	IP addresses are owned by the complete host on Linux, not by
1143	particular interfaces. Only for more complex setups like load-
1144	balancing, does this behaviour cause problems.
1145
1146	arp_filter for the interface will be enabled if at least one of
1147	conf/{all,interface}/arp_filter is set to TRUE,
1148	it will be disabled otherwise
1149
1150arp_announce - INTEGER
1151	Define different restriction levels for announcing the local
1152	source IP address from IP packets in ARP requests sent on
1153	interface:
1154	0 - (default) Use any local address, configured on any interface
1155	1 - Try to avoid local addresses that are not in the target's
1156	subnet for this interface. This mode is useful when target
1157	hosts reachable via this interface require the source IP
1158	address in ARP requests to be part of their logical network
1159	configured on the receiving interface. When we generate the
1160	request we will check all our subnets that include the
1161	target IP and will preserve the source address if it is from
1162	such subnet. If there is no such subnet we select source
1163	address according to the rules for level 2.
1164	2 - Always use the best local address for this target.
1165	In this mode we ignore the source address in the IP packet
1166	and try to select local address that we prefer for talks with
1167	the target host. Such local address is selected by looking
1168	for primary IP addresses on all our subnets on the outgoing
1169	interface that include the target IP address. If no suitable
1170	local address is found we select the first local address
1171	we have on the outgoing interface or on all other interfaces,
1172	with the hope we will receive reply for our request and
1173	even sometimes no matter the source IP address we announce.
1174
1175	The max value from conf/{all,interface}/arp_announce is used.
1176
1177	Increasing the restriction level gives more chance for
1178	receiving answer from the resolved target while decreasing
1179	the level announces more valid sender's information.
1180
1181arp_ignore - INTEGER
1182	Define different modes for sending replies in response to
1183	received ARP requests that resolve local target IP addresses:
1184	0 - (default): reply for any local target IP address, configured
1185	on any interface
1186	1 - reply only if the target IP address is local address
1187	configured on the incoming interface
1188	2 - reply only if the target IP address is local address
1189	configured on the incoming interface and both with the
1190	sender's IP address are part from same subnet on this interface
1191	3 - do not reply for local addresses configured with scope host,
1192	only resolutions for global and link addresses are replied
1193	4-7 - reserved
1194	8 - do not reply for all local addresses
1195
1196	The max value from conf/{all,interface}/arp_ignore is used
1197	when ARP request is received on the {interface}
1198
1199arp_notify - BOOLEAN
1200	Define mode for notification of address and device changes.
1201	0 - (default): do nothing
1202	1 - Generate gratuitous arp requests when device is brought up
1203	    or hardware address changes.
1204
1205arp_accept - BOOLEAN
1206	Define behavior for gratuitous ARP frames who's IP is not
1207	already present in the ARP table:
1208	0 - don't create new entries in the ARP table
1209	1 - create new entries in the ARP table
1210
1211	Both replies and requests type gratuitous arp will trigger the
1212	ARP table to be updated, if this setting is on.
1213
1214	If the ARP table already contains the IP address of the
1215	gratuitous arp frame, the arp table will be updated regardless
1216	if this setting is on or off.
1217
1218mcast_solicit - INTEGER
1219	The maximum number of multicast probes in INCOMPLETE state,
1220	when the associated hardware address is unknown.  Defaults
1221	to 3.
1222
1223ucast_solicit - INTEGER
1224	The maximum number of unicast probes in PROBE state, when
1225	the hardware address is being reconfirmed.  Defaults to 3.
1226
1227app_solicit - INTEGER
1228	The maximum number of probes to send to the user space ARP daemon
1229	via netlink before dropping back to multicast probes (see
1230	mcast_resolicit).  Defaults to 0.
1231
1232mcast_resolicit - INTEGER
1233	The maximum number of multicast probes after unicast and
1234	app probes in PROBE state.  Defaults to 0.
1235
1236disable_policy - BOOLEAN
1237	Disable IPSEC policy (SPD) for this interface
1238
1239disable_xfrm - BOOLEAN
1240	Disable IPSEC encryption on this interface, whatever the policy
1241
1242igmpv2_unsolicited_report_interval - INTEGER
1243	The interval in milliseconds in which the next unsolicited
1244	IGMPv1 or IGMPv2 report retransmit will take place.
1245	Default: 10000 (10 seconds)
1246
1247igmpv3_unsolicited_report_interval - INTEGER
1248	The interval in milliseconds in which the next unsolicited
1249	IGMPv3 report retransmit will take place.
1250	Default: 1000 (1 seconds)
1251
1252promote_secondaries - BOOLEAN
1253	When a primary IP address is removed from this interface
1254	promote a corresponding secondary IP address instead of
1255	removing all the corresponding secondary IP addresses.
1256
1257drop_unicast_in_l2_multicast - BOOLEAN
1258	Drop any unicast IP packets that are received in link-layer
1259	multicast (or broadcast) frames.
1260	This behavior (for multicast) is actually a SHOULD in RFC
1261	1122, but is disabled by default for compatibility reasons.
1262	Default: off (0)
1263
1264drop_gratuitous_arp - BOOLEAN
1265	Drop all gratuitous ARP frames, for example if there's a known
1266	good ARP proxy on the network and such frames need not be used
1267	(or in the case of 802.11, must not be used to prevent attacks.)
1268	Default: off (0)
1269
1270
1271tag - INTEGER
1272	Allows you to write a number, which can be used as required.
1273	Default value is 0.
1274
1275xfrm4_gc_thresh - INTEGER
1276	The threshold at which we will start garbage collecting for IPv4
1277	destination cache entries.  At twice this value the system will
1278	refuse new allocations. The value must be set below the flowcache
1279	limit (4096 * number of online cpus) to take effect.
1280
1281igmp_link_local_mcast_reports - BOOLEAN
1282	Enable IGMP reports for link local multicast groups in the
1283	224.0.0.X range.
1284	Default TRUE
1285
1286Alexey Kuznetsov.
1287kuznet@ms2.inr.ac.ru
1288
1289Updated by:
1290Andi Kleen
1291ak@muc.de
1292Nicolas Delon
1293delon.nicolas@wanadoo.fr
1294
1295
1296
1297
1298/proc/sys/net/ipv6/* Variables:
1299
1300IPv6 has no global variables such as tcp_*.  tcp_* settings under ipv4/ also
1301apply to IPv6 [XXX?].
1302
1303bindv6only - BOOLEAN
1304	Default value for IPV6_V6ONLY socket option,
1305	which restricts use of the IPv6 socket to IPv6 communication
1306	only.
1307		TRUE: disable IPv4-mapped address feature
1308		FALSE: enable IPv4-mapped address feature
1309
1310	Default: FALSE (as specified in RFC3493)
1311
1312flowlabel_consistency - BOOLEAN
1313	Protect the consistency (and unicity) of flow label.
1314	You have to disable it to use IPV6_FL_F_REFLECT flag on the
1315	flow label manager.
1316	TRUE: enabled
1317	FALSE: disabled
1318	Default: TRUE
1319
1320auto_flowlabels - INTEGER
1321	Automatically generate flow labels based on a flow hash of the
1322	packet. This allows intermediate devices, such as routers, to
1323	identify packet flows for mechanisms like Equal Cost Multipath
1324	Routing (see RFC 6438).
1325	0: automatic flow labels are completely disabled
1326	1: automatic flow labels are enabled by default, they can be
1327	   disabled on a per socket basis using the IPV6_AUTOFLOWLABEL
1328	   socket option
1329	2: automatic flow labels are allowed, they may be enabled on a
1330	   per socket basis using the IPV6_AUTOFLOWLABEL socket option
1331	3: automatic flow labels are enabled and enforced, they cannot
1332	   be disabled by the socket option
1333	Default: 1
1334
1335flowlabel_state_ranges - BOOLEAN
1336	Split the flow label number space into two ranges. 0-0x7FFFF is
1337	reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF
1338	is reserved for stateless flow labels as described in RFC6437.
1339	TRUE: enabled
1340	FALSE: disabled
1341	Default: true
1342
1343anycast_src_echo_reply - BOOLEAN
1344	Controls the use of anycast addresses as source addresses for ICMPv6
1345	echo reply
1346	TRUE:  enabled
1347	FALSE: disabled
1348	Default: FALSE
1349
1350idgen_delay - INTEGER
1351	Controls the delay in seconds after which time to retry
1352	privacy stable address generation if a DAD conflict is
1353	detected.
1354	Default: 1 (as specified in RFC7217)
1355
1356idgen_retries - INTEGER
1357	Controls the number of retries to generate a stable privacy
1358	address if a DAD conflict is detected.
1359	Default: 3 (as specified in RFC7217)
1360
1361mld_qrv - INTEGER
1362	Controls the MLD query robustness variable (see RFC3810 9.1).
1363	Default: 2 (as specified by RFC3810 9.1)
1364	Minimum: 1 (as specified by RFC6636 4.5)
1365
1366IPv6 Fragmentation:
1367
1368ip6frag_high_thresh - INTEGER
1369	Maximum memory used to reassemble IPv6 fragments. When
1370	ip6frag_high_thresh bytes of memory is allocated for this purpose,
1371	the fragment handler will toss packets until ip6frag_low_thresh
1372	is reached.
1373
1374ip6frag_low_thresh - INTEGER
1375	See ip6frag_high_thresh
1376
1377ip6frag_time - INTEGER
1378	Time in seconds to keep an IPv6 fragment in memory.
1379
1380conf/default/*:
1381	Change the interface-specific default settings.
1382
1383
1384conf/all/*:
1385	Change all the interface-specific settings.
1386
1387	[XXX:  Other special features than forwarding?]
1388
1389conf/all/forwarding - BOOLEAN
1390	Enable global IPv6 forwarding between all interfaces.
1391
1392	IPv4 and IPv6 work differently here; e.g. netfilter must be used
1393	to control which interfaces may forward packets and which not.
1394
1395	This also sets all interfaces' Host/Router setting
1396	'forwarding' to the specified value.  See below for details.
1397
1398	This referred to as global forwarding.
1399
1400proxy_ndp - BOOLEAN
1401	Do proxy ndp.
1402
1403fwmark_reflect - BOOLEAN
1404	Controls the fwmark of kernel-generated IPv6 reply packets that are not
1405	associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
1406	If unset, these packets have a fwmark of zero. If set, they have the
1407	fwmark of the packet they are replying to.
1408	Default: 0
1409
1410conf/interface/*:
1411	Change special settings per interface.
1412
1413	The functional behaviour for certain settings is different
1414	depending on whether local forwarding is enabled or not.
1415
1416accept_ra - INTEGER
1417	Accept Router Advertisements; autoconfigure using them.
1418
1419	It also determines whether or not to transmit Router
1420	Solicitations. If and only if the functional setting is to
1421	accept Router Advertisements, Router Solicitations will be
1422	transmitted.
1423
1424	Possible values are:
1425		0 Do not accept Router Advertisements.
1426		1 Accept Router Advertisements if forwarding is disabled.
1427		2 Overrule forwarding behaviour. Accept Router Advertisements
1428		  even if forwarding is enabled.
1429
1430	Functional default: enabled if local forwarding is disabled.
1431			    disabled if local forwarding is enabled.
1432
1433accept_ra_defrtr - BOOLEAN
1434	Learn default router in Router Advertisement.
1435
1436	Functional default: enabled if accept_ra is enabled.
1437			    disabled if accept_ra is disabled.
1438
1439accept_ra_from_local - BOOLEAN
1440	Accept RA with source-address that is found on local machine
1441        if the RA is otherwise proper and able to be accepted.
1442        Default is to NOT accept these as it may be an un-intended
1443        network loop.
1444
1445	Functional default:
1446           enabled if accept_ra_from_local is enabled
1447               on a specific interface.
1448	   disabled if accept_ra_from_local is disabled
1449               on a specific interface.
1450
1451accept_ra_min_hop_limit - INTEGER
1452	Minimum hop limit Information in Router Advertisement.
1453
1454	Hop limit Information in Router Advertisement less than this
1455	variable shall be ignored.
1456
1457	Default: 1
1458
1459accept_ra_pinfo - BOOLEAN
1460	Learn Prefix Information in Router Advertisement.
1461
1462	Functional default: enabled if accept_ra is enabled.
1463			    disabled if accept_ra is disabled.
1464
1465accept_ra_rt_info_max_plen - INTEGER
1466	Maximum prefix length of Route Information in RA.
1467
1468	Route Information w/ prefix larger than or equal to this
1469	variable shall be ignored.
1470
1471	Functional default: 0 if accept_ra_rtr_pref is enabled.
1472			    -1 if accept_ra_rtr_pref is disabled.
1473
1474accept_ra_rtr_pref - BOOLEAN
1475	Accept Router Preference in RA.
1476
1477	Functional default: enabled if accept_ra is enabled.
1478			    disabled if accept_ra is disabled.
1479
1480accept_ra_mtu - BOOLEAN
1481	Apply the MTU value specified in RA option 5 (RFC4861). If
1482	disabled, the MTU specified in the RA will be ignored.
1483
1484	Functional default: enabled if accept_ra is enabled.
1485			    disabled if accept_ra is disabled.
1486
1487accept_redirects - BOOLEAN
1488	Accept Redirects.
1489
1490	Functional default: enabled if local forwarding is disabled.
1491			    disabled if local forwarding is enabled.
1492
1493accept_source_route - INTEGER
1494	Accept source routing (routing extension header).
1495
1496	>= 0: Accept only routing header type 2.
1497	< 0: Do not accept routing header.
1498
1499	Default: 0
1500
1501autoconf - BOOLEAN
1502	Autoconfigure addresses using Prefix Information in Router
1503	Advertisements.
1504
1505	Functional default: enabled if accept_ra_pinfo is enabled.
1506			    disabled if accept_ra_pinfo is disabled.
1507
1508dad_transmits - INTEGER
1509	The amount of Duplicate Address Detection probes to send.
1510	Default: 1
1511
1512forwarding - INTEGER
1513	Configure interface-specific Host/Router behaviour.
1514
1515	Note: It is recommended to have the same setting on all
1516	interfaces; mixed router/host scenarios are rather uncommon.
1517
1518	Possible values are:
1519		0 Forwarding disabled
1520		1 Forwarding enabled
1521
1522	FALSE (0):
1523
1524	By default, Host behaviour is assumed.  This means:
1525
1526	1. IsRouter flag is not set in Neighbour Advertisements.
1527	2. If accept_ra is TRUE (default), transmit Router
1528	   Solicitations.
1529	3. If accept_ra is TRUE (default), accept Router
1530	   Advertisements (and do autoconfiguration).
1531	4. If accept_redirects is TRUE (default), accept Redirects.
1532
1533	TRUE (1):
1534
1535	If local forwarding is enabled, Router behaviour is assumed.
1536	This means exactly the reverse from the above:
1537
1538	1. IsRouter flag is set in Neighbour Advertisements.
1539	2. Router Solicitations are not sent unless accept_ra is 2.
1540	3. Router Advertisements are ignored unless accept_ra is 2.
1541	4. Redirects are ignored.
1542
1543	Default: 0 (disabled) if global forwarding is disabled (default),
1544		 otherwise 1 (enabled).
1545
1546hop_limit - INTEGER
1547	Default Hop Limit to set.
1548	Default: 64
1549
1550mtu - INTEGER
1551	Default Maximum Transfer Unit
1552	Default: 1280 (IPv6 required minimum)
1553
1554ip_nonlocal_bind - BOOLEAN
1555	If set, allows processes to bind() to non-local IPv6 addresses,
1556	which can be quite useful - but may break some applications.
1557	Default: 0
1558
1559router_probe_interval - INTEGER
1560	Minimum interval (in seconds) between Router Probing described
1561	in RFC4191.
1562
1563	Default: 60
1564
1565router_solicitation_delay - INTEGER
1566	Number of seconds to wait after interface is brought up
1567	before sending Router Solicitations.
1568	Default: 1
1569
1570router_solicitation_interval - INTEGER
1571	Number of seconds to wait between Router Solicitations.
1572	Default: 4
1573
1574router_solicitations - INTEGER
1575	Number of Router Solicitations to send until assuming no
1576	routers are present.
1577	Default: 3
1578
1579use_oif_addrs_only - BOOLEAN
1580	When enabled, the candidate source addresses for destinations
1581	routed via this interface are restricted to the set of addresses
1582	configured on this interface (vis. RFC 6724, section 4).
1583
1584	Default: false
1585
1586use_tempaddr - INTEGER
1587	Preference for Privacy Extensions (RFC3041).
1588	  <= 0 : disable Privacy Extensions
1589	  == 1 : enable Privacy Extensions, but prefer public
1590	         addresses over temporary addresses.
1591	  >  1 : enable Privacy Extensions and prefer temporary
1592	         addresses over public addresses.
1593	Default:  0 (for most devices)
1594		 -1 (for point-to-point devices and loopback devices)
1595
1596temp_valid_lft - INTEGER
1597	valid lifetime (in seconds) for temporary addresses.
1598	Default: 604800 (7 days)
1599
1600temp_prefered_lft - INTEGER
1601	Preferred lifetime (in seconds) for temporary addresses.
1602	Default: 86400 (1 day)
1603
1604keep_addr_on_down - INTEGER
1605	Keep all IPv6 addresses on an interface down event. If set static
1606	global addresses with no expiration time are not flushed.
1607	  >0 : enabled
1608	   0 : system default
1609	  <0 : disabled
1610
1611	Default: 0 (addresses are removed)
1612
1613max_desync_factor - INTEGER
1614	Maximum value for DESYNC_FACTOR, which is a random value
1615	that ensures that clients don't synchronize with each
1616	other and generate new addresses at exactly the same time.
1617	value is in seconds.
1618	Default: 600
1619
1620regen_max_retry - INTEGER
1621	Number of attempts before give up attempting to generate
1622	valid temporary addresses.
1623	Default: 5
1624
1625max_addresses - INTEGER
1626	Maximum number of autoconfigured addresses per interface.  Setting
1627	to zero disables the limitation.  It is not recommended to set this
1628	value too large (or to zero) because it would be an easy way to
1629	crash the kernel by allowing too many addresses to be created.
1630	Default: 16
1631
1632disable_ipv6 - BOOLEAN
1633	Disable IPv6 operation.  If accept_dad is set to 2, this value
1634	will be dynamically set to TRUE if DAD fails for the link-local
1635	address.
1636	Default: FALSE (enable IPv6 operation)
1637
1638	When this value is changed from 1 to 0 (IPv6 is being enabled),
1639	it will dynamically create a link-local address on the given
1640	interface and start Duplicate Address Detection, if necessary.
1641
1642	When this value is changed from 0 to 1 (IPv6 is being disabled),
1643	it will dynamically delete all address on the given interface.
1644
1645accept_dad - INTEGER
1646	Whether to accept DAD (Duplicate Address Detection).
1647	0: Disable DAD
1648	1: Enable DAD (default)
1649	2: Enable DAD, and disable IPv6 operation if MAC-based duplicate
1650	   link-local address has been found.
1651
1652force_tllao - BOOLEAN
1653	Enable sending the target link-layer address option even when
1654	responding to a unicast neighbor solicitation.
1655	Default: FALSE
1656
1657	Quoting from RFC 2461, section 4.4, Target link-layer address:
1658
1659	"The option MUST be included for multicast solicitations in order to
1660	avoid infinite Neighbor Solicitation "recursion" when the peer node
1661	does not have a cache entry to return a Neighbor Advertisements
1662	message.  When responding to unicast solicitations, the option can be
1663	omitted since the sender of the solicitation has the correct link-
1664	layer address; otherwise it would not have be able to send the unicast
1665	solicitation in the first place. However, including the link-layer
1666	address in this case adds little overhead and eliminates a potential
1667	race condition where the sender deletes the cached link-layer address
1668	prior to receiving a response to a previous solicitation."
1669
1670ndisc_notify - BOOLEAN
1671	Define mode for notification of address and device changes.
1672	0 - (default): do nothing
1673	1 - Generate unsolicited neighbour advertisements when device is brought
1674	    up or hardware address changes.
1675
1676mldv1_unsolicited_report_interval - INTEGER
1677	The interval in milliseconds in which the next unsolicited
1678	MLDv1 report retransmit will take place.
1679	Default: 10000 (10 seconds)
1680
1681mldv2_unsolicited_report_interval - INTEGER
1682	The interval in milliseconds in which the next unsolicited
1683	MLDv2 report retransmit will take place.
1684	Default: 1000 (1 second)
1685
1686force_mld_version - INTEGER
1687	0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
1688	1 - Enforce to use MLD version 1
1689	2 - Enforce to use MLD version 2
1690
1691suppress_frag_ndisc - INTEGER
1692	Control RFC 6980 (Security Implications of IPv6 Fragmentation
1693	with IPv6 Neighbor Discovery) behavior:
1694	1 - (default) discard fragmented neighbor discovery packets
1695	0 - allow fragmented neighbor discovery packets
1696
1697optimistic_dad - BOOLEAN
1698	Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
1699		0: disabled (default)
1700		1: enabled
1701
1702use_optimistic - BOOLEAN
1703	If enabled, do not classify optimistic addresses as deprecated during
1704	source address selection.  Preferred addresses will still be chosen
1705	before optimistic addresses, subject to other ranking in the source
1706	address selection algorithm.
1707		0: disabled (default)
1708		1: enabled
1709
1710stable_secret - IPv6 address
1711	This IPv6 address will be used as a secret to generate IPv6
1712	addresses for link-local addresses and autoconfigured
1713	ones. All addresses generated after setting this secret will
1714	be stable privacy ones by default. This can be changed via the
1715	addrgenmode ip-link. conf/default/stable_secret is used as the
1716	secret for the namespace, the interface specific ones can
1717	overwrite that. Writes to conf/all/stable_secret are refused.
1718
1719	It is recommended to generate this secret during installation
1720	of a system and keep it stable after that.
1721
1722	By default the stable secret is unset.
1723
1724drop_unicast_in_l2_multicast - BOOLEAN
1725	Drop any unicast IPv6 packets that are received in link-layer
1726	multicast (or broadcast) frames.
1727
1728	By default this is turned off.
1729
1730drop_unsolicited_na - BOOLEAN
1731	Drop all unsolicited neighbor advertisements, for example if there's
1732	a known good NA proxy on the network and such frames need not be used
1733	(or in the case of 802.11, must not be used to prevent attacks.)
1734
1735	By default this is turned off.
1736
1737enhanced_dad - BOOLEAN
1738	Include a nonce option in the IPv6 neighbor solicitation messages used for
1739	duplicate address detection per RFC7527. A received DAD NS will only signal
1740	a duplicate address if the nonce is different. This avoids any false
1741	detection of duplicates due to loopback of the NS messages that we send.
1742	The nonce option will be sent on an interface unless both of
1743	conf/{all,interface}/enhanced_dad are set to FALSE.
1744	Default: TRUE
1745
1746icmp/*:
1747ratelimit - INTEGER
1748	Limit the maximal rates for sending ICMPv6 packets.
1749	0 to disable any limiting,
1750	otherwise the minimal space between responses in milliseconds.
1751	Default: 1000
1752
1753xfrm6_gc_thresh - INTEGER
1754	The threshold at which we will start garbage collecting for IPv6
1755	destination cache entries.  At twice this value the system will
1756	refuse new allocations. The value must be set below the flowcache
1757	limit (4096 * number of online cpus) to take effect.
1758
1759
1760IPv6 Update by:
1761Pekka Savola <pekkas@netcore.fi>
1762YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1763
1764
1765/proc/sys/net/bridge/* Variables:
1766
1767bridge-nf-call-arptables - BOOLEAN
1768	1 : pass bridged ARP traffic to arptables' FORWARD chain.
1769	0 : disable this.
1770	Default: 1
1771
1772bridge-nf-call-iptables - BOOLEAN
1773	1 : pass bridged IPv4 traffic to iptables' chains.
1774	0 : disable this.
1775	Default: 1
1776
1777bridge-nf-call-ip6tables - BOOLEAN
1778	1 : pass bridged IPv6 traffic to ip6tables' chains.
1779	0 : disable this.
1780	Default: 1
1781
1782bridge-nf-filter-vlan-tagged - BOOLEAN
1783	1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1784	0 : disable this.
1785	Default: 0
1786
1787bridge-nf-filter-pppoe-tagged - BOOLEAN
1788	1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1789	0 : disable this.
1790	Default: 0
1791
1792bridge-nf-pass-vlan-input-dev - BOOLEAN
1793	1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
1794	interface on the bridge and set the netfilter input device to the vlan.
1795	This allows use of e.g. "iptables -i br0.1" and makes the REDIRECT
1796	target work with vlan-on-top-of-bridge interfaces.  When no matching
1797	vlan interface is found, or this switch is off, the input device is
1798	set to the bridge interface.
1799	0: disable bridge netfilter vlan interface lookup.
1800	Default: 0
1801
1802proc/sys/net/sctp/* Variables:
1803
1804addip_enable - BOOLEAN
1805	Enable or disable extension of  Dynamic Address Reconfiguration
1806	(ADD-IP) functionality specified in RFC5061.  This extension provides
1807	the ability to dynamically add and remove new addresses for the SCTP
1808	associations.
1809
1810	1: Enable extension.
1811
1812	0: Disable extension.
1813
1814	Default: 0
1815
1816pf_enable - INTEGER
1817	Enable or disable pf (pf is short for potentially failed) state. A value
1818	of pf_retrans > path_max_retrans also disables pf state. That is, one of
1819	both pf_enable and pf_retrans > path_max_retrans can disable pf state.
1820	Since pf_retrans and path_max_retrans can be changed by userspace
1821	application, sometimes user expects to disable pf state by the value of
1822	pf_retrans > path_max_retrans, but occasionally the value of pf_retrans
1823	or path_max_retrans is changed by the user application, this pf state is
1824	enabled. As such, it is necessary to add this to dynamically enable
1825	and disable pf state. See:
1826	https://datatracker.ietf.org/doc/draft-ietf-tsvwg-sctp-failover for
1827	details.
1828
1829	1: Enable pf.
1830
1831	0: Disable pf.
1832
1833	Default: 1
1834
1835addip_noauth_enable - BOOLEAN
1836	Dynamic Address Reconfiguration (ADD-IP) requires the use of
1837	authentication to protect the operations of adding or removing new
1838	addresses.  This requirement is mandated so that unauthorized hosts
1839	would not be able to hijack associations.  However, older
1840	implementations may not have implemented this requirement while
1841	allowing the ADD-IP extension.  For reasons of interoperability,
1842	we provide this variable to control the enforcement of the
1843	authentication requirement.
1844
1845	1: Allow ADD-IP extension to be used without authentication.  This
1846	   should only be set in a closed environment for interoperability
1847	   with older implementations.
1848
1849	0: Enforce the authentication requirement
1850
1851	Default: 0
1852
1853auth_enable - BOOLEAN
1854	Enable or disable Authenticated Chunks extension.  This extension
1855	provides the ability to send and receive authenticated chunks and is
1856	required for secure operation of Dynamic Address Reconfiguration
1857	(ADD-IP) extension.
1858
1859	1: Enable this extension.
1860	0: Disable this extension.
1861
1862	Default: 0
1863
1864prsctp_enable - BOOLEAN
1865	Enable or disable the Partial Reliability extension (RFC3758) which
1866	is used to notify peers that a given DATA should no longer be expected.
1867
1868	1: Enable extension
1869	0: Disable
1870
1871	Default: 1
1872
1873max_burst - INTEGER
1874	The limit of the number of new packets that can be initially sent.  It
1875	controls how bursty the generated traffic can be.
1876
1877	Default: 4
1878
1879association_max_retrans - INTEGER
1880	Set the maximum number for retransmissions that an association can
1881	attempt deciding that the remote end is unreachable.  If this value
1882	is exceeded, the association is terminated.
1883
1884	Default: 10
1885
1886max_init_retransmits - INTEGER
1887	The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
1888	that an association will attempt before declaring the destination
1889	unreachable and terminating.
1890
1891	Default: 8
1892
1893path_max_retrans - INTEGER
1894	The maximum number of retransmissions that will be attempted on a given
1895	path.  Once this threshold is exceeded, the path is considered
1896	unreachable, and new traffic will use a different path when the
1897	association is multihomed.
1898
1899	Default: 5
1900
1901pf_retrans - INTEGER
1902	The number of retransmissions that will be attempted on a given path
1903	before traffic is redirected to an alternate transport (should one
1904	exist).  Note this is distinct from path_max_retrans, as a path that
1905	passes the pf_retrans threshold can still be used.  Its only
1906	deprioritized when a transmission path is selected by the stack.  This
1907	setting is primarily used to enable fast failover mechanisms without
1908	having to reduce path_max_retrans to a very low value.  See:
1909	http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
1910	for details.  Note also that a value of pf_retrans > path_max_retrans
1911	disables this feature. Since both pf_retrans and path_max_retrans can
1912	be changed by userspace application, a variable pf_enable is used to
1913	disable pf state.
1914
1915	Default: 0
1916
1917rto_initial - INTEGER
1918	The initial round trip timeout value in milliseconds that will be used
1919	in calculating round trip times.  This is the initial time interval
1920	for retransmissions.
1921
1922	Default: 3000
1923
1924rto_max - INTEGER
1925	The maximum value (in milliseconds) of the round trip timeout.  This
1926	is the largest time interval that can elapse between retransmissions.
1927
1928	Default: 60000
1929
1930rto_min - INTEGER
1931	The minimum value (in milliseconds) of the round trip timeout.  This
1932	is the smallest time interval the can elapse between retransmissions.
1933
1934	Default: 1000
1935
1936hb_interval - INTEGER
1937	The interval (in milliseconds) between HEARTBEAT chunks.  These chunks
1938	are sent at the specified interval on idle paths to probe the state of
1939	a given path between 2 associations.
1940
1941	Default: 30000
1942
1943sack_timeout - INTEGER
1944	The amount of time (in milliseconds) that the implementation will wait
1945	to send a SACK.
1946
1947	Default: 200
1948
1949valid_cookie_life - INTEGER
1950	The default lifetime of the SCTP cookie (in milliseconds).  The cookie
1951	is used during association establishment.
1952
1953	Default: 60000
1954
1955cookie_preserve_enable - BOOLEAN
1956	Enable or disable the ability to extend the lifetime of the SCTP cookie
1957	that is used during the establishment phase of SCTP association
1958
1959	1: Enable cookie lifetime extension.
1960	0: Disable
1961
1962	Default: 1
1963
1964cookie_hmac_alg - STRING
1965	Select the hmac algorithm used when generating the cookie value sent by
1966	a listening sctp socket to a connecting client in the INIT-ACK chunk.
1967	Valid values are:
1968	* md5
1969	* sha1
1970	* none
1971	Ability to assign md5 or sha1 as the selected alg is predicated on the
1972	configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and
1973	CONFIG_CRYPTO_SHA1).
1974
1975	Default: Dependent on configuration.  MD5 if available, else SHA1 if
1976	available, else none.
1977
1978rcvbuf_policy - INTEGER
1979	Determines if the receive buffer is attributed to the socket or to
1980	association.   SCTP supports the capability to create multiple
1981	associations on a single socket.  When using this capability, it is
1982	possible that a single stalled association that's buffering a lot
1983	of data may block other associations from delivering their data by
1984	consuming all of the receive buffer space.  To work around this,
1985	the rcvbuf_policy could be set to attribute the receiver buffer space
1986	to each association instead of the socket.  This prevents the described
1987	blocking.
1988
1989	1: rcvbuf space is per association
1990	0: rcvbuf space is per socket
1991
1992	Default: 0
1993
1994sndbuf_policy - INTEGER
1995	Similar to rcvbuf_policy above, this applies to send buffer space.
1996
1997	1: Send buffer is tracked per association
1998	0: Send buffer is tracked per socket.
1999
2000	Default: 0
2001
2002sctp_mem - vector of 3 INTEGERs: min, pressure, max
2003	Number of pages allowed for queueing by all SCTP sockets.
2004
2005	min: Below this number of pages SCTP is not bothered about its
2006	memory appetite. When amount of memory allocated by SCTP exceeds
2007	this number, SCTP starts to moderate memory usage.
2008
2009	pressure: This value was introduced to follow format of tcp_mem.
2010
2011	max: Number of pages allowed for queueing by all SCTP sockets.
2012
2013	Default is calculated at boot time from amount of available memory.
2014
2015sctp_rmem - vector of 3 INTEGERs: min, default, max
2016	Only the first value ("min") is used, "default" and "max" are
2017	ignored.
2018
2019	min: Minimal size of receive buffer used by SCTP socket.
2020	It is guaranteed to each SCTP socket (but not association) even
2021	under moderate memory pressure.
2022
2023	Default: 1 page
2024
2025sctp_wmem  - vector of 3 INTEGERs: min, default, max
2026	Currently this tunable has no effect.
2027
2028addr_scope_policy - INTEGER
2029	Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
2030
2031	0   - Disable IPv4 address scoping
2032	1   - Enable IPv4 address scoping
2033	2   - Follow draft but allow IPv4 private addresses
2034	3   - Follow draft but allow IPv4 link local addresses
2035
2036	Default: 1
2037
2038
2039/proc/sys/net/core/*
2040	Please see: Documentation/sysctl/net.txt for descriptions of these entries.
2041
2042
2043/proc/sys/net/unix/*
2044max_dgram_qlen - INTEGER
2045	The maximum length of dgram socket receive queue
2046
2047	Default: 10
2048
2049
2050UNDOCUMENTED:
2051
2052/proc/sys/net/irda/*
2053	fast_poll_increase FIXME
2054	warn_noreply_time FIXME
2055	discovery_slots FIXME
2056	slot_timeout FIXME
2057	max_baud_rate FIXME
2058	discovery_timeout FIXME
2059	lap_keepalive_time FIXME
2060	max_noreply_time FIXME
2061	max_tx_data_size FIXME
2062	max_tx_window FIXME
2063	min_tx_turn_time FIXME