About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Documentation / networking / ip-sysctl.txt




Custom Search

Based on kernel version 4.0. Page generated on 2015-04-14 21:25 EST.

1	/proc/sys/net/ipv4/* Variables:
2	
3	ip_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	
13	ip_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	
18	ip_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	
43	min_pmtu - INTEGER
44		default 552 - minimum discovered Path MTU
45	
46	ip_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	
59	fwmark_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	
66	route/max_size - INTEGER
67		Maximum number of routes allowed in the kernel.  Increase
68		this when using large numbers of interfaces and/or routes.
69		From linux kernel 3.6 onwards, this is deprecated for ipv4
70		as route cache is no longer used.
71	
72	neigh/default/gc_thresh1 - INTEGER
73		Minimum number of entries to keep.  Garbage collector will not
74		purge entries if there are fewer than this number.
75		Default: 128
76	
77	neigh/default/gc_thresh2 - INTEGER
78		Threshold when garbage collector becomes more aggressive about
79		purging entries. Entries older than 5 seconds will be cleared
80		when over this number.
81		Default: 512
82	
83	neigh/default/gc_thresh3 - INTEGER
84		Maximum number of neighbor entries allowed.  Increase this
85		when using large numbers of interfaces and when communicating
86		with large numbers of directly-connected peers.
87		Default: 1024
88	
89	neigh/default/unres_qlen_bytes - INTEGER
90		The maximum number of bytes which may be used by packets
91		queued for each	unresolved address by other network layers.
92		(added in linux 3.3)
93		Setting negative value is meaningless and will return error.
94		Default: 65536 Bytes(64KB)
95	
96	neigh/default/unres_qlen - INTEGER
97		The maximum number of packets which may be queued for each
98		unresolved address by other network layers.
99		(deprecated in linux 3.3) : use unres_qlen_bytes instead.
100		Prior to linux 3.3, the default value is 3 which may cause
101		unexpected packet loss. The current default value is calculated
102		according to default value of unres_qlen_bytes and true size of
103		packet.
104		Default: 31
105	
106	mtu_expires - INTEGER
107		Time, in seconds, that cached PMTU information is kept.
108	
109	min_adv_mss - INTEGER
110		The advertised MSS depends on the first hop route MTU, but will
111		never be lower than this setting.
112	
113	IP Fragmentation:
114	
115	ipfrag_high_thresh - INTEGER
116		Maximum memory used to reassemble IP fragments. When
117		ipfrag_high_thresh bytes of memory is allocated for this purpose,
118		the fragment handler will toss packets until ipfrag_low_thresh
119		is reached. This also serves as a maximum limit to namespaces
120		different from the initial one.
121	
122	ipfrag_low_thresh - INTEGER
123		Maximum memory used to reassemble IP fragments before the kernel
124		begins to remove incomplete fragment queues to free up resources.
125		The kernel still accepts new fragments for defragmentation.
126	
127	ipfrag_time - INTEGER
128		Time in seconds to keep an IP fragment in memory.
129	
130	ipfrag_max_dist - INTEGER
131		ipfrag_max_dist is a non-negative integer value which defines the
132		maximum "disorder" which is allowed among fragments which share a
133		common IP source address. Note that reordering of packets is
134		not unusual, but if a large number of fragments arrive from a source
135		IP address while a particular fragment queue remains incomplete, it
136		probably indicates that one or more fragments belonging to that queue
137		have been lost. When ipfrag_max_dist is positive, an additional check
138		is done on fragments before they are added to a reassembly queue - if
139		ipfrag_max_dist (or more) fragments have arrived from a particular IP
140		address between additions to any IP fragment queue using that source
141		address, it's presumed that one or more fragments in the queue are
142		lost. The existing fragment queue will be dropped, and a new one
143		started. An ipfrag_max_dist value of zero disables this check.
144	
145		Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
146		result in unnecessarily dropping fragment queues when normal
147		reordering of packets occurs, which could lead to poor application
148		performance. Using a very large value, e.g. 50000, increases the
149		likelihood of incorrectly reassembling IP fragments that originate
150		from different IP datagrams, which could result in data corruption.
151		Default: 64
152	
153	INET peer storage:
154	
155	inet_peer_threshold - INTEGER
156		The approximate size of the storage.  Starting from this threshold
157		entries will be thrown aggressively.  This threshold also determines
158		entries' time-to-live and time intervals between garbage collection
159		passes.  More entries, less time-to-live, less GC interval.
160	
161	inet_peer_minttl - INTEGER
162		Minimum time-to-live of entries.  Should be enough to cover fragment
163		time-to-live on the reassembling side.  This minimum time-to-live  is
164		guaranteed if the pool size is less than inet_peer_threshold.
165		Measured in seconds.
166	
167	inet_peer_maxttl - INTEGER
168		Maximum time-to-live of entries.  Unused entries will expire after
169		this period of time if there is no memory pressure on the pool (i.e.
170		when the number of entries in the pool is very small).
171		Measured in seconds.
172	
173	TCP variables:
174	
175	somaxconn - INTEGER
176		Limit of socket listen() backlog, known in userspace as SOMAXCONN.
177		Defaults to 128.  See also tcp_max_syn_backlog for additional tuning
178		for TCP sockets.
179	
180	tcp_abort_on_overflow - BOOLEAN
181		If listening service is too slow to accept new connections,
182		reset them. Default state is FALSE. It means that if overflow
183		occurred due to a burst, connection will recover. Enable this
184		option _only_ if you are really sure that listening daemon
185		cannot be tuned to accept connections faster. Enabling this
186		option can harm clients of your server.
187	
188	tcp_adv_win_scale - INTEGER
189		Count buffering overhead as bytes/2^tcp_adv_win_scale
190		(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
191		if it is <= 0.
192		Possible values are [-31, 31], inclusive.
193		Default: 1
194	
195	tcp_allowed_congestion_control - STRING
196		Show/set the congestion control choices available to non-privileged
197		processes. The list is a subset of those listed in
198		tcp_available_congestion_control.
199		Default is "reno" and the default setting (tcp_congestion_control).
200	
201	tcp_app_win - INTEGER
202		Reserve max(window/2^tcp_app_win, mss) of window for application
203		buffer. Value 0 is special, it means that nothing is reserved.
204		Default: 31
205	
206	tcp_autocorking - BOOLEAN
207		Enable TCP auto corking :
208		When applications do consecutive small write()/sendmsg() system calls,
209		we try to coalesce these small writes as much as possible, to lower
210		total amount of sent packets. This is done if at least one prior
211		packet for the flow is waiting in Qdisc queues or device transmit
212		queue. Applications can still use TCP_CORK for optimal behavior
213		when they know how/when to uncork their sockets.
214		Default : 1
215	
216	tcp_available_congestion_control - STRING
217		Shows the available congestion control choices that are registered.
218		More congestion control algorithms may be available as modules,
219		but not loaded.
220	
221	tcp_base_mss - INTEGER
222		The initial value of search_low to be used by the packetization layer
223		Path MTU discovery (MTU probing).  If MTU probing is enabled,
224		this is the initial MSS used by the connection.
225	
226	tcp_congestion_control - STRING
227		Set the congestion control algorithm to be used for new
228		connections. The algorithm "reno" is always available, but
229		additional choices may be available based on kernel configuration.
230		Default is set as part of kernel configuration.
231		For passive connections, the listener congestion control choice
232		is inherited.
233		[see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
234	
235	tcp_dsack - BOOLEAN
236		Allows TCP to send "duplicate" SACKs.
237	
238	tcp_early_retrans - INTEGER
239		Enable Early Retransmit (ER), per RFC 5827. ER lowers the threshold
240		for triggering fast retransmit when the amount of outstanding data is
241		small and when no previously unsent data can be transmitted (such
242		that limited transmit could be used). Also controls the use of
243		Tail loss probe (TLP) that converts RTOs occurring due to tail
244		losses into fast recovery (draft-dukkipati-tcpm-tcp-loss-probe-01).
245		Possible values:
246			0 disables ER
247			1 enables ER
248			2 enables ER but delays fast recovery and fast retransmit
249			  by a fourth of RTT. This mitigates connection falsely
250			  recovers when network has a small degree of reordering
251			  (less than 3 packets).
252			3 enables delayed ER and TLP.
253			4 enables TLP only.
254		Default: 3
255	
256	tcp_ecn - INTEGER
257		Control use of Explicit Congestion Notification (ECN) by TCP.
258		ECN is used only when both ends of the TCP connection indicate
259		support for it.  This feature is useful in avoiding losses due
260		to congestion by allowing supporting routers to signal
261		congestion before having to drop packets.
262		Possible values are:
263			0 Disable ECN.  Neither initiate nor accept ECN.
264			1 Enable ECN when requested by incoming connections and
265			  also request ECN on outgoing connection attempts.
266			2 Enable ECN when requested by incoming connections
267			  but do not request ECN on outgoing connections.
268		Default: 2
269	
270	tcp_fack - BOOLEAN
271		Enable FACK congestion avoidance and fast retransmission.
272		The value is not used, if tcp_sack is not enabled.
273	
274	tcp_fin_timeout - INTEGER
275		The length of time an orphaned (no longer referenced by any
276		application) connection will remain in the FIN_WAIT_2 state
277		before it is aborted at the local end.  While a perfectly
278		valid "receive only" state for an un-orphaned connection, an
279		orphaned connection in FIN_WAIT_2 state could otherwise wait
280		forever for the remote to close its end of the connection.
281		Cf. tcp_max_orphans
282		Default: 60 seconds
283	
284	tcp_frto - INTEGER
285		Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
286		F-RTO is an enhanced recovery algorithm for TCP retransmission
287		timeouts.  It is particularly beneficial in networks where the
288		RTT fluctuates (e.g., wireless). F-RTO is sender-side only
289		modification. It does not require any support from the peer.
290	
291		By default it's enabled with a non-zero value. 0 disables F-RTO.
292	
293	tcp_invalid_ratelimit - INTEGER
294		Limit the maximal rate for sending duplicate acknowledgments
295		in response to incoming TCP packets that are for an existing
296		connection but that are invalid due to any of these reasons:
297	
298		  (a) out-of-window sequence number,
299		  (b) out-of-window acknowledgment number, or
300		  (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
301	
302		This can help mitigate simple "ack loop" DoS attacks, wherein
303		a buggy or malicious middlebox or man-in-the-middle can
304		rewrite TCP header fields in manner that causes each endpoint
305		to think that the other is sending invalid TCP segments, thus
306		causing each side to send an unterminating stream of duplicate
307		acknowledgments for invalid segments.
308	
309		Using 0 disables rate-limiting of dupacks in response to
310		invalid segments; otherwise this value specifies the minimal
311		space between sending such dupacks, in milliseconds.
312	
313		Default: 500 (milliseconds).
314	
315	tcp_keepalive_time - INTEGER
316		How often TCP sends out keepalive messages when keepalive is enabled.
317		Default: 2hours.
318	
319	tcp_keepalive_probes - INTEGER
320		How many keepalive probes TCP sends out, until it decides that the
321		connection is broken. Default value: 9.
322	
323	tcp_keepalive_intvl - INTEGER
324		How frequently the probes are send out. Multiplied by
325		tcp_keepalive_probes it is time to kill not responding connection,
326		after probes started. Default value: 75sec i.e. connection
327		will be aborted after ~11 minutes of retries.
328	
329	tcp_low_latency - BOOLEAN
330		If set, the TCP stack makes decisions that prefer lower
331		latency as opposed to higher throughput.  By default, this
332		option is not set meaning that higher throughput is preferred.
333		An example of an application where this default should be
334		changed would be a Beowulf compute cluster.
335		Default: 0
336	
337	tcp_max_orphans - INTEGER
338		Maximal number of TCP sockets not attached to any user file handle,
339		held by system.	If this number is exceeded orphaned connections are
340		reset immediately and warning is printed. This limit exists
341		only to prevent simple DoS attacks, you _must_ not rely on this
342		or lower the limit artificially, but rather increase it
343		(probably, after increasing installed memory),
344		if network conditions require more than default value,
345		and tune network services to linger and kill such states
346		more aggressively. Let me to remind again: each orphan eats
347		up to ~64K of unswappable memory.
348	
349	tcp_max_syn_backlog - INTEGER
350		Maximal number of remembered connection requests, which have not
351		received an acknowledgment from connecting client.
352		The minimal value is 128 for low memory machines, and it will
353		increase in proportion to the memory of machine.
354		If server suffers from overload, try increasing this number.
355	
356	tcp_max_tw_buckets - INTEGER
357		Maximal number of timewait sockets held by system simultaneously.
358		If this number is exceeded time-wait socket is immediately destroyed
359		and warning is printed. This limit exists only to prevent
360		simple DoS attacks, you _must_ not lower the limit artificially,
361		but rather increase it (probably, after increasing installed memory),
362		if network conditions require more than default value.
363	
364	tcp_mem - vector of 3 INTEGERs: min, pressure, max
365		min: below this number of pages TCP is not bothered about its
366		memory appetite.
367	
368		pressure: when amount of memory allocated by TCP exceeds this number
369		of pages, TCP moderates its memory consumption and enters memory
370		pressure mode, which is exited when memory consumption falls
371		under "min".
372	
373		max: number of pages allowed for queueing by all TCP sockets.
374	
375		Defaults are calculated at boot time from amount of available
376		memory.
377	
378	tcp_moderate_rcvbuf - BOOLEAN
379		If set, TCP performs receive buffer auto-tuning, attempting to
380		automatically size the buffer (no greater than tcp_rmem[2]) to
381		match the size required by the path for full throughput.  Enabled by
382		default.
383	
384	tcp_mtu_probing - INTEGER
385		Controls TCP Packetization-Layer Path MTU Discovery.  Takes three
386		values:
387		  0 - Disabled
388		  1 - Disabled by default, enabled when an ICMP black hole detected
389		  2 - Always enabled, use initial MSS of tcp_base_mss.
390	
391	tcp_no_metrics_save - BOOLEAN
392		By default, TCP saves various connection metrics in the route cache
393		when the connection closes, so that connections established in the
394		near future can use these to set initial conditions.  Usually, this
395		increases overall performance, but may sometimes cause performance
396		degradation.  If set, TCP will not cache metrics on closing
397		connections.
398	
399	tcp_orphan_retries - INTEGER
400		This value influences the timeout of a locally closed TCP connection,
401		when RTO retransmissions remain unacknowledged.
402		See tcp_retries2 for more details.
403	
404		The default value is 8.
405		If your machine is a loaded WEB server,
406		you should think about lowering this value, such sockets
407		may consume significant resources. Cf. tcp_max_orphans.
408	
409	tcp_reordering - INTEGER
410		Initial reordering level of packets in a TCP stream.
411		TCP stack can then dynamically adjust flow reordering level
412		between this initial value and tcp_max_reordering
413		Default: 3
414	
415	tcp_max_reordering - INTEGER
416		Maximal reordering level of packets in a TCP stream.
417		300 is a fairly conservative value, but you might increase it
418		if paths are using per packet load balancing (like bonding rr mode)
419		Default: 300
420	
421	tcp_retrans_collapse - BOOLEAN
422		Bug-to-bug compatibility with some broken printers.
423		On retransmit try to send bigger packets to work around bugs in
424		certain TCP stacks.
425	
426	tcp_retries1 - INTEGER
427		This value influences the time, after which TCP decides, that
428		something is wrong due to unacknowledged RTO retransmissions,
429		and reports this suspicion to the network layer.
430		See tcp_retries2 for more details.
431	
432		RFC 1122 recommends at least 3 retransmissions, which is the
433		default.
434	
435	tcp_retries2 - INTEGER
436		This value influences the timeout of an alive TCP connection,
437		when RTO retransmissions remain unacknowledged.
438		Given a value of N, a hypothetical TCP connection following
439		exponential backoff with an initial RTO of TCP_RTO_MIN would
440		retransmit N times before killing the connection at the (N+1)th RTO.
441	
442		The default value of 15 yields a hypothetical timeout of 924.6
443		seconds and is a lower bound for the effective timeout.
444		TCP will effectively time out at the first RTO which exceeds the
445		hypothetical timeout.
446	
447		RFC 1122 recommends at least 100 seconds for the timeout,
448		which corresponds to a value of at least 8.
449	
450	tcp_rfc1337 - BOOLEAN
451		If set, the TCP stack behaves conforming to RFC1337. If unset,
452		we are not conforming to RFC, but prevent TCP TIME_WAIT
453		assassination.
454		Default: 0
455	
456	tcp_rmem - vector of 3 INTEGERs: min, default, max
457		min: Minimal size of receive buffer used by TCP sockets.
458		It is guaranteed to each TCP socket, even under moderate memory
459		pressure.
460		Default: 1 page
461	
462		default: initial size of receive buffer used by TCP sockets.
463		This value overrides net.core.rmem_default used by other protocols.
464		Default: 87380 bytes. This value results in window of 65535 with
465		default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
466		less for default tcp_app_win. See below about these variables.
467	
468		max: maximal size of receive buffer allowed for automatically
469		selected receiver buffers for TCP socket. This value does not override
470		net.core.rmem_max.  Calling setsockopt() with SO_RCVBUF disables
471		automatic tuning of that socket's receive buffer size, in which
472		case this value is ignored.
473		Default: between 87380B and 6MB, depending on RAM size.
474	
475	tcp_sack - BOOLEAN
476		Enable select acknowledgments (SACKS).
477	
478	tcp_slow_start_after_idle - BOOLEAN
479		If set, provide RFC2861 behavior and time out the congestion
480		window after an idle period.  An idle period is defined at
481		the current RTO.  If unset, the congestion window will not
482		be timed out after an idle period.
483		Default: 1
484	
485	tcp_stdurg - BOOLEAN
486		Use the Host requirements interpretation of the TCP urgent pointer field.
487		Most hosts use the older BSD interpretation, so if you turn this on
488		Linux might not communicate correctly with them.
489		Default: FALSE
490	
491	tcp_synack_retries - INTEGER
492		Number of times SYNACKs for a passive TCP connection attempt will
493		be retransmitted. Should not be higher than 255. Default value
494		is 5, which corresponds to 31seconds till the last retransmission
495		with the current initial RTO of 1second. With this the final timeout
496		for a passive TCP connection will happen after 63seconds.
497	
498	tcp_syncookies - BOOLEAN
499		Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
500		Send out syncookies when the syn backlog queue of a socket
501		overflows. This is to prevent against the common 'SYN flood attack'
502		Default: 1
503	
504		Note, that syncookies is fallback facility.
505		It MUST NOT be used to help highly loaded servers to stand
506		against legal connection rate. If you see SYN flood warnings
507		in your logs, but investigation	shows that they occur
508		because of overload with legal connections, you should tune
509		another parameters until this warning disappear.
510		See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
511	
512		syncookies seriously violate TCP protocol, do not allow
513		to use TCP extensions, can result in serious degradation
514		of some services (f.e. SMTP relaying), visible not by you,
515		but your clients and relays, contacting you. While you see
516		SYN flood warnings in logs not being really flooded, your server
517		is seriously misconfigured.
518	
519		If you want to test which effects syncookies have to your
520		network connections you can set this knob to 2 to enable
521		unconditionally generation of syncookies.
522	
523	tcp_fastopen - INTEGER
524		Enable TCP Fast Open feature (draft-ietf-tcpm-fastopen) to send data
525		in the opening SYN packet. To use this feature, the client application
526		must use sendmsg() or sendto() with MSG_FASTOPEN flag rather than
527		connect() to perform a TCP handshake automatically.
528	
529		The values (bitmap) are
530		1: Enables sending data in the opening SYN on the client w/ MSG_FASTOPEN.
531		2: Enables TCP Fast Open on the server side, i.e., allowing data in
532		   a SYN packet to be accepted and passed to the application before
533		   3-way hand shake finishes.
534		4: Send data in the opening SYN regardless of cookie availability and
535		   without a cookie option.
536		0x100: Accept SYN data w/o validating the cookie.
537		0x200: Accept data-in-SYN w/o any cookie option present.
538		0x400/0x800: Enable Fast Open on all listeners regardless of the
539		   TCP_FASTOPEN socket option. The two different flags designate two
540		   different ways of setting max_qlen without the TCP_FASTOPEN socket
541		   option.
542	
543		Default: 1
544	
545		Note that the client & server side Fast Open flags (1 and 2
546		respectively) must be also enabled before the rest of flags can take
547		effect.
548	
549		See include/net/tcp.h and the code for more details.
550	
551	tcp_syn_retries - INTEGER
552		Number of times initial SYNs for an active TCP connection attempt
553		will be retransmitted. Should not be higher than 255. Default value
554		is 6, which corresponds to 63seconds till the last retransmission
555		with the current initial RTO of 1second. With this the final timeout
556		for an active TCP connection attempt will happen after 127seconds.
557	
558	tcp_timestamps - BOOLEAN
559		Enable timestamps as defined in RFC1323.
560	
561	tcp_min_tso_segs - INTEGER
562		Minimal number of segments per TSO frame.
563		Since linux-3.12, TCP does an automatic sizing of TSO frames,
564		depending on flow rate, instead of filling 64Kbytes packets.
565		For specific usages, it's possible to force TCP to build big
566		TSO frames. Note that TCP stack might split too big TSO packets
567		if available window is too small.
568		Default: 2
569	
570	tcp_tso_win_divisor - INTEGER
571		This allows control over what percentage of the congestion window
572		can be consumed by a single TSO frame.
573		The setting of this parameter is a choice between burstiness and
574		building larger TSO frames.
575		Default: 3
576	
577	tcp_tw_recycle - BOOLEAN
578		Enable fast recycling TIME-WAIT sockets. Default value is 0.
579		It should not be changed without advice/request of technical
580		experts.
581	
582	tcp_tw_reuse - BOOLEAN
583		Allow to reuse TIME-WAIT sockets for new connections when it is
584		safe from protocol viewpoint. Default value is 0.
585		It should not be changed without advice/request of technical
586		experts.
587	
588	tcp_window_scaling - BOOLEAN
589		Enable window scaling as defined in RFC1323.
590	
591	tcp_wmem - vector of 3 INTEGERs: min, default, max
592		min: Amount of memory reserved for send buffers for TCP sockets.
593		Each TCP socket has rights to use it due to fact of its birth.
594		Default: 1 page
595	
596		default: initial size of send buffer used by TCP sockets.  This
597		value overrides net.core.wmem_default used by other protocols.
598		It is usually lower than net.core.wmem_default.
599		Default: 16K
600	
601		max: Maximal amount of memory allowed for automatically tuned
602		send buffers for TCP sockets. This value does not override
603		net.core.wmem_max.  Calling setsockopt() with SO_SNDBUF disables
604		automatic tuning of that socket's send buffer size, in which case
605		this value is ignored.
606		Default: between 64K and 4MB, depending on RAM size.
607	
608	tcp_notsent_lowat - UNSIGNED INTEGER
609		A TCP socket can control the amount of unsent bytes in its write queue,
610		thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
611		reports POLLOUT events if the amount of unsent bytes is below a per
612		socket value, and if the write queue is not full. sendmsg() will
613		also not add new buffers if the limit is hit.
614	
615		This global variable controls the amount of unsent data for
616		sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
617		to the global variable has immediate effect.
618	
619		Default: UINT_MAX (0xFFFFFFFF)
620	
621	tcp_workaround_signed_windows - BOOLEAN
622		If set, assume no receipt of a window scaling option means the
623		remote TCP is broken and treats the window as a signed quantity.
624		If unset, assume the remote TCP is not broken even if we do
625		not receive a window scaling option from them.
626		Default: 0
627	
628	tcp_thin_linear_timeouts - BOOLEAN
629		Enable dynamic triggering of linear timeouts for thin streams.
630		If set, a check is performed upon retransmission by timeout to
631		determine if the stream is thin (less than 4 packets in flight).
632		As long as the stream is found to be thin, up to 6 linear
633		timeouts may be performed before exponential backoff mode is
634		initiated. This improves retransmission latency for
635		non-aggressive thin streams, often found to be time-dependent.
636		For more information on thin streams, see
637		Documentation/networking/tcp-thin.txt
638		Default: 0
639	
640	tcp_thin_dupack - BOOLEAN
641		Enable dynamic triggering of retransmissions after one dupACK
642		for thin streams. If set, a check is performed upon reception
643		of a dupACK to determine if the stream is thin (less than 4
644		packets in flight). As long as the stream is found to be thin,
645		data is retransmitted on the first received dupACK. This
646		improves retransmission latency for non-aggressive thin
647		streams, often found to be time-dependent.
648		For more information on thin streams, see
649		Documentation/networking/tcp-thin.txt
650		Default: 0
651	
652	tcp_limit_output_bytes - INTEGER
653		Controls TCP Small Queue limit per tcp socket.
654		TCP bulk sender tends to increase packets in flight until it
655		gets losses notifications. With SNDBUF autotuning, this can
656		result in a large amount of packets queued in qdisc/device
657		on the local machine, hurting latency of other flows, for
658		typical pfifo_fast qdiscs.
659		tcp_limit_output_bytes limits the number of bytes on qdisc
660		or device to reduce artificial RTT/cwnd and reduce bufferbloat.
661		Default: 131072
662	
663	tcp_challenge_ack_limit - INTEGER
664		Limits number of Challenge ACK sent per second, as recommended
665		in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
666		Default: 100
667	
668	UDP variables:
669	
670	udp_mem - vector of 3 INTEGERs: min, pressure, max
671		Number of pages allowed for queueing by all UDP sockets.
672	
673		min: Below this number of pages UDP is not bothered about its
674		memory appetite. When amount of memory allocated by UDP exceeds
675		this number, UDP starts to moderate memory usage.
676	
677		pressure: This value was introduced to follow format of tcp_mem.
678	
679		max: Number of pages allowed for queueing by all UDP sockets.
680	
681		Default is calculated at boot time from amount of available memory.
682	
683	udp_rmem_min - INTEGER
684		Minimal size of receive buffer used by UDP sockets in moderation.
685		Each UDP socket is able to use the size for receiving data, even if
686		total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
687		Default: 1 page
688	
689	udp_wmem_min - INTEGER
690		Minimal size of send buffer used by UDP sockets in moderation.
691		Each UDP socket is able to use the size for sending data, even if
692		total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
693		Default: 1 page
694	
695	CIPSOv4 Variables:
696	
697	cipso_cache_enable - BOOLEAN
698		If set, enable additions to and lookups from the CIPSO label mapping
699		cache.  If unset, additions are ignored and lookups always result in a
700		miss.  However, regardless of the setting the cache is still
701		invalidated when required when means you can safely toggle this on and
702		off and the cache will always be "safe".
703		Default: 1
704	
705	cipso_cache_bucket_size - INTEGER
706		The CIPSO label cache consists of a fixed size hash table with each
707		hash bucket containing a number of cache entries.  This variable limits
708		the number of entries in each hash bucket; the larger the value the
709		more CIPSO label mappings that can be cached.  When the number of
710		entries in a given hash bucket reaches this limit adding new entries
711		causes the oldest entry in the bucket to be removed to make room.
712		Default: 10
713	
714	cipso_rbm_optfmt - BOOLEAN
715		Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
716		the CIPSO draft specification (see Documentation/netlabel for details).
717		This means that when set the CIPSO tag will be padded with empty
718		categories in order to make the packet data 32-bit aligned.
719		Default: 0
720	
721	cipso_rbm_structvalid - BOOLEAN
722		If set, do a very strict check of the CIPSO option when
723		ip_options_compile() is called.  If unset, relax the checks done during
724		ip_options_compile().  Either way is "safe" as errors are caught else
725		where in the CIPSO processing code but setting this to 0 (False) should
726		result in less work (i.e. it should be faster) but could cause problems
727		with other implementations that require strict checking.
728		Default: 0
729	
730	IP Variables:
731	
732	ip_local_port_range - 2 INTEGERS
733		Defines the local port range that is used by TCP and UDP to
734		choose the local port. The first number is the first, the
735		second the last local port number. The default values are
736		32768 and 61000 respectively.
737	
738	ip_local_reserved_ports - list of comma separated ranges
739		Specify the ports which are reserved for known third-party
740		applications. These ports will not be used by automatic port
741		assignments (e.g. when calling connect() or bind() with port
742		number 0). Explicit port allocation behavior is unchanged.
743	
744		The format used for both input and output is a comma separated
745		list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
746		10). Writing to the file will clear all previously reserved
747		ports and update the current list with the one given in the
748		input.
749	
750		Note that ip_local_port_range and ip_local_reserved_ports
751		settings are independent and both are considered by the kernel
752		when determining which ports are available for automatic port
753		assignments.
754	
755		You can reserve ports which are not in the current
756		ip_local_port_range, e.g.:
757	
758		$ cat /proc/sys/net/ipv4/ip_local_port_range
759		32000	61000
760		$ cat /proc/sys/net/ipv4/ip_local_reserved_ports
761		8080,9148
762	
763		although this is redundant. However such a setting is useful
764		if later the port range is changed to a value that will
765		include the reserved ports.
766	
767		Default: Empty
768	
769	ip_nonlocal_bind - BOOLEAN
770		If set, allows processes to bind() to non-local IP addresses,
771		which can be quite useful - but may break some applications.
772		Default: 0
773	
774	ip_dynaddr - BOOLEAN
775		If set non-zero, enables support for dynamic addresses.
776		If set to a non-zero value larger than 1, a kernel log
777		message will be printed when dynamic address rewriting
778		occurs.
779		Default: 0
780	
781	ip_early_demux - BOOLEAN
782		Optimize input packet processing down to one demux for
783		certain kinds of local sockets.  Currently we only do this
784		for established TCP sockets.
785	
786		It may add an additional cost for pure routing workloads that
787		reduces overall throughput, in such case you should disable it.
788		Default: 1
789	
790	icmp_echo_ignore_all - BOOLEAN
791		If set non-zero, then the kernel will ignore all ICMP ECHO
792		requests sent to it.
793		Default: 0
794	
795	icmp_echo_ignore_broadcasts - BOOLEAN
796		If set non-zero, then the kernel will ignore all ICMP ECHO and
797		TIMESTAMP requests sent to it via broadcast/multicast.
798		Default: 1
799	
800	icmp_ratelimit - INTEGER
801		Limit the maximal rates for sending ICMP packets whose type matches
802		icmp_ratemask (see below) to specific targets.
803		0 to disable any limiting,
804		otherwise the minimal space between responses in milliseconds.
805		Note that another sysctl, icmp_msgs_per_sec limits the number
806		of ICMP packets	sent on all targets.
807		Default: 1000
808	
809	icmp_msgs_per_sec - INTEGER
810		Limit maximal number of ICMP packets sent per second from this host.
811		Only messages whose type matches icmp_ratemask (see below) are
812		controlled by this limit.
813		Default: 1000
814	
815	icmp_msgs_burst - INTEGER
816		icmp_msgs_per_sec controls number of ICMP packets sent per second,
817		while icmp_msgs_burst controls the burst size of these packets.
818		Default: 50
819	
820	icmp_ratemask - INTEGER
821		Mask made of ICMP types for which rates are being limited.
822		Significant bits: IHGFEDCBA9876543210
823		Default mask:     0000001100000011000 (6168)
824	
825		Bit definitions (see include/linux/icmp.h):
826			0 Echo Reply
827			3 Destination Unreachable *
828			4 Source Quench *
829			5 Redirect
830			8 Echo Request
831			B Time Exceeded *
832			C Parameter Problem *
833			D Timestamp Request
834			E Timestamp Reply
835			F Info Request
836			G Info Reply
837			H Address Mask Request
838			I Address Mask Reply
839	
840		* These are rate limited by default (see default mask above)
841	
842	icmp_ignore_bogus_error_responses - BOOLEAN
843		Some routers violate RFC1122 by sending bogus responses to broadcast
844		frames.  Such violations are normally logged via a kernel warning.
845		If this is set to TRUE, the kernel will not give such warnings, which
846		will avoid log file clutter.
847		Default: 1
848	
849	icmp_errors_use_inbound_ifaddr - BOOLEAN
850	
851		If zero, icmp error messages are sent with the primary address of
852		the exiting interface.
853	
854		If non-zero, the message will be sent with the primary address of
855		the interface that received the packet that caused the icmp error.
856		This is the behaviour network many administrators will expect from
857		a router. And it can make debugging complicated network layouts
858		much easier.
859	
860		Note that if no primary address exists for the interface selected,
861		then the primary address of the first non-loopback interface that
862		has one will be used regardless of this setting.
863	
864		Default: 0
865	
866	igmp_max_memberships - INTEGER
867		Change the maximum number of multicast groups we can subscribe to.
868		Default: 20
869	
870		Theoretical maximum value is bounded by having to send a membership
871		report in a single datagram (i.e. the report can't span multiple
872		datagrams, or risk confusing the switch and leaving groups you don't
873		intend to).
874	
875		The number of supported groups 'M' is bounded by the number of group
876		report entries you can fit into a single datagram of 65535 bytes.
877	
878		M = 65536-sizeof (ip header)/(sizeof(Group record))
879	
880		Group records are variable length, with a minimum of 12 bytes.
881		So net.ipv4.igmp_max_memberships should not be set higher than:
882	
883		(65536-24) / 12 = 5459
884	
885		The value 5459 assumes no IP header options, so in practice
886		this number may be lower.
887	
888		conf/interface/*  changes special settings per interface (where
889		"interface" is the name of your network interface)
890	
891		conf/all/*	  is special, changes the settings for all interfaces
892	
893	igmp_qrv - INTEGER
894		 Controls the IGMP query robustness variable (see RFC2236 8.1).
895		 Default: 2 (as specified by RFC2236 8.1)
896		 Minimum: 1 (as specified by RFC6636 4.5)
897	
898	log_martians - BOOLEAN
899		Log packets with impossible addresses to kernel log.
900		log_martians for the interface will be enabled if at least one of
901		conf/{all,interface}/log_martians is set to TRUE,
902		it will be disabled otherwise
903	
904	accept_redirects - BOOLEAN
905		Accept ICMP redirect messages.
906		accept_redirects for the interface will be enabled if:
907		- both conf/{all,interface}/accept_redirects are TRUE in the case
908		  forwarding for the interface is enabled
909		or
910		- at least one of conf/{all,interface}/accept_redirects is TRUE in the
911		  case forwarding for the interface is disabled
912		accept_redirects for the interface will be disabled otherwise
913		default TRUE (host)
914			FALSE (router)
915	
916	forwarding - BOOLEAN
917		Enable IP forwarding on this interface.
918	
919	mc_forwarding - BOOLEAN
920		Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
921		and a multicast routing daemon is required.
922		conf/all/mc_forwarding must also be set to TRUE to enable multicast
923		routing	for the interface
924	
925	medium_id - INTEGER
926		Integer value used to differentiate the devices by the medium they
927		are attached to. Two devices can have different id values when
928		the broadcast packets are received only on one of them.
929		The default value 0 means that the device is the only interface
930		to its medium, value of -1 means that medium is not known.
931	
932		Currently, it is used to change the proxy_arp behavior:
933		the proxy_arp feature is enabled for packets forwarded between
934		two devices attached to different media.
935	
936	proxy_arp - BOOLEAN
937		Do proxy arp.
938		proxy_arp for the interface will be enabled if at least one of
939		conf/{all,interface}/proxy_arp is set to TRUE,
940		it will be disabled otherwise
941	
942	proxy_arp_pvlan - BOOLEAN
943		Private VLAN proxy arp.
944		Basically allow proxy arp replies back to the same interface
945		(from which the ARP request/solicitation was received).
946	
947		This is done to support (ethernet) switch features, like RFC
948		3069, where the individual ports are NOT allowed to
949		communicate with each other, but they are allowed to talk to
950		the upstream router.  As described in RFC 3069, it is possible
951		to allow these hosts to communicate through the upstream
952		router by proxy_arp'ing. Don't need to be used together with
953		proxy_arp.
954	
955		This technology is known by different names:
956		  In RFC 3069 it is called VLAN Aggregation.
957		  Cisco and Allied Telesyn call it Private VLAN.
958		  Hewlett-Packard call it Source-Port filtering or port-isolation.
959		  Ericsson call it MAC-Forced Forwarding (RFC Draft).
960	
961	shared_media - BOOLEAN
962		Send(router) or accept(host) RFC1620 shared media redirects.
963		Overrides ip_secure_redirects.
964		shared_media for the interface will be enabled if at least one of
965		conf/{all,interface}/shared_media is set to TRUE,
966		it will be disabled otherwise
967		default TRUE
968	
969	secure_redirects - BOOLEAN
970		Accept ICMP redirect messages only for gateways,
971		listed in default gateway list.
972		secure_redirects for the interface will be enabled if at least one of
973		conf/{all,interface}/secure_redirects is set to TRUE,
974		it will be disabled otherwise
975		default TRUE
976	
977	send_redirects - BOOLEAN
978		Send redirects, if router.
979		send_redirects for the interface will be enabled if at least one of
980		conf/{all,interface}/send_redirects is set to TRUE,
981		it will be disabled otherwise
982		Default: TRUE
983	
984	bootp_relay - BOOLEAN
985		Accept packets with source address 0.b.c.d destined
986		not to this host as local ones. It is supposed, that
987		BOOTP relay daemon will catch and forward such packets.
988		conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
989		for the interface
990		default FALSE
991		Not Implemented Yet.
992	
993	accept_source_route - BOOLEAN
994		Accept packets with SRR option.
995		conf/all/accept_source_route must also be set to TRUE to accept packets
996		with SRR option on the interface
997		default TRUE (router)
998			FALSE (host)
999	
1000	accept_local - BOOLEAN
1001		Accept packets with local source addresses. In combination with
1002		suitable routing, this can be used to direct packets between two
1003		local interfaces over the wire and have them accepted properly.
1004		default FALSE
1005	
1006	route_localnet - BOOLEAN
1007		Do not consider loopback addresses as martian source or destination
1008		while routing. This enables the use of 127/8 for local routing purposes.
1009		default FALSE
1010	
1011	rp_filter - INTEGER
1012		0 - No source validation.
1013		1 - Strict mode as defined in RFC3704 Strict Reverse Path
1014		    Each incoming packet is tested against the FIB and if the interface
1015		    is not the best reverse path the packet check will fail.
1016		    By default failed packets are discarded.
1017		2 - Loose mode as defined in RFC3704 Loose Reverse Path
1018		    Each incoming packet's source address is also tested against the FIB
1019		    and if the source address is not reachable via any interface
1020		    the packet check will fail.
1021	
1022		Current recommended practice in RFC3704 is to enable strict mode
1023		to prevent IP spoofing from DDos attacks. If using asymmetric routing
1024		or other complicated routing, then loose mode is recommended.
1025	
1026		The max value from conf/{all,interface}/rp_filter is used
1027		when doing source validation on the {interface}.
1028	
1029		Default value is 0. Note that some distributions enable it
1030		in startup scripts.
1031	
1032	arp_filter - BOOLEAN
1033		1 - Allows you to have multiple network interfaces on the same
1034		subnet, and have the ARPs for each interface be answered
1035		based on whether or not the kernel would route a packet from
1036		the ARP'd IP out that interface (therefore you must use source
1037		based routing for this to work). In other words it allows control
1038		of which cards (usually 1) will respond to an arp request.
1039	
1040		0 - (default) The kernel can respond to arp requests with addresses
1041		from other interfaces. This may seem wrong but it usually makes
1042		sense, because it increases the chance of successful communication.
1043		IP addresses are owned by the complete host on Linux, not by
1044		particular interfaces. Only for more complex setups like load-
1045		balancing, does this behaviour cause problems.
1046	
1047		arp_filter for the interface will be enabled if at least one of
1048		conf/{all,interface}/arp_filter is set to TRUE,
1049		it will be disabled otherwise
1050	
1051	arp_announce - INTEGER
1052		Define different restriction levels for announcing the local
1053		source IP address from IP packets in ARP requests sent on
1054		interface:
1055		0 - (default) Use any local address, configured on any interface
1056		1 - Try to avoid local addresses that are not in the target's
1057		subnet for this interface. This mode is useful when target
1058		hosts reachable via this interface require the source IP
1059		address in ARP requests to be part of their logical network
1060		configured on the receiving interface. When we generate the
1061		request we will check all our subnets that include the
1062		target IP and will preserve the source address if it is from
1063		such subnet. If there is no such subnet we select source
1064		address according to the rules for level 2.
1065		2 - Always use the best local address for this target.
1066		In this mode we ignore the source address in the IP packet
1067		and try to select local address that we prefer for talks with
1068		the target host. Such local address is selected by looking
1069		for primary IP addresses on all our subnets on the outgoing
1070		interface that include the target IP address. If no suitable
1071		local address is found we select the first local address
1072		we have on the outgoing interface or on all other interfaces,
1073		with the hope we will receive reply for our request and
1074		even sometimes no matter the source IP address we announce.
1075	
1076		The max value from conf/{all,interface}/arp_announce is used.
1077	
1078		Increasing the restriction level gives more chance for
1079		receiving answer from the resolved target while decreasing
1080		the level announces more valid sender's information.
1081	
1082	arp_ignore - INTEGER
1083		Define different modes for sending replies in response to
1084		received ARP requests that resolve local target IP addresses:
1085		0 - (default): reply for any local target IP address, configured
1086		on any interface
1087		1 - reply only if the target IP address is local address
1088		configured on the incoming interface
1089		2 - reply only if the target IP address is local address
1090		configured on the incoming interface and both with the
1091		sender's IP address are part from same subnet on this interface
1092		3 - do not reply for local addresses configured with scope host,
1093		only resolutions for global and link addresses are replied
1094		4-7 - reserved
1095		8 - do not reply for all local addresses
1096	
1097		The max value from conf/{all,interface}/arp_ignore is used
1098		when ARP request is received on the {interface}
1099	
1100	arp_notify - BOOLEAN
1101		Define mode for notification of address and device changes.
1102		0 - (default): do nothing
1103		1 - Generate gratuitous arp requests when device is brought up
1104		    or hardware address changes.
1105	
1106	arp_accept - BOOLEAN
1107		Define behavior for gratuitous ARP frames who's IP is not
1108		already present in the ARP table:
1109		0 - don't create new entries in the ARP table
1110		1 - create new entries in the ARP table
1111	
1112		Both replies and requests type gratuitous arp will trigger the
1113		ARP table to be updated, if this setting is on.
1114	
1115		If the ARP table already contains the IP address of the
1116		gratuitous arp frame, the arp table will be updated regardless
1117		if this setting is on or off.
1118	
1119	
1120	app_solicit - INTEGER
1121		The maximum number of probes to send to the user space ARP daemon
1122		via netlink before dropping back to multicast probes (see
1123		mcast_solicit).  Defaults to 0.
1124	
1125	disable_policy - BOOLEAN
1126		Disable IPSEC policy (SPD) for this interface
1127	
1128	disable_xfrm - BOOLEAN
1129		Disable IPSEC encryption on this interface, whatever the policy
1130	
1131	igmpv2_unsolicited_report_interval - INTEGER
1132		The interval in milliseconds in which the next unsolicited
1133		IGMPv1 or IGMPv2 report retransmit will take place.
1134		Default: 10000 (10 seconds)
1135	
1136	igmpv3_unsolicited_report_interval - INTEGER
1137		The interval in milliseconds in which the next unsolicited
1138		IGMPv3 report retransmit will take place.
1139		Default: 1000 (1 seconds)
1140	
1141	promote_secondaries - BOOLEAN
1142		When a primary IP address is removed from this interface
1143		promote a corresponding secondary IP address instead of
1144		removing all the corresponding secondary IP addresses.
1145	
1146	
1147	tag - INTEGER
1148		Allows you to write a number, which can be used as required.
1149		Default value is 0.
1150	
1151	Alexey Kuznetsov.
1152	kuznet@ms2.inr.ac.ru
1153	
1154	Updated by:
1155	Andi Kleen
1156	ak@muc.de
1157	Nicolas Delon
1158	delon.nicolas@wanadoo.fr
1159	
1160	
1161	
1162	
1163	/proc/sys/net/ipv6/* Variables:
1164	
1165	IPv6 has no global variables such as tcp_*.  tcp_* settings under ipv4/ also
1166	apply to IPv6 [XXX?].
1167	
1168	bindv6only - BOOLEAN
1169		Default value for IPV6_V6ONLY socket option,
1170		which restricts use of the IPv6 socket to IPv6 communication
1171		only.
1172			TRUE: disable IPv4-mapped address feature
1173			FALSE: enable IPv4-mapped address feature
1174	
1175		Default: FALSE (as specified in RFC3493)
1176	
1177	flowlabel_consistency - BOOLEAN
1178		Protect the consistency (and unicity) of flow label.
1179		You have to disable it to use IPV6_FL_F_REFLECT flag on the
1180		flow label manager.
1181		TRUE: enabled
1182		FALSE: disabled
1183		Default: TRUE
1184	
1185	auto_flowlabels - BOOLEAN
1186		Automatically generate flow labels based based on a flow hash
1187		of the packet. This allows intermediate devices, such as routers,
1188		to idenfify packet flows for mechanisms like Equal Cost Multipath
1189		Routing (see RFC 6438).
1190		TRUE: enabled
1191		FALSE: disabled
1192		Default: false
1193	
1194	anycast_src_echo_reply - BOOLEAN
1195		Controls the use of anycast addresses as source addresses for ICMPv6
1196		echo reply
1197		TRUE:  enabled
1198		FALSE: disabled
1199		Default: FALSE
1200	
1201	mld_qrv - INTEGER
1202		Controls the MLD query robustness variable (see RFC3810 9.1).
1203		Default: 2 (as specified by RFC3810 9.1)
1204		Minimum: 1 (as specified by RFC6636 4.5)
1205	
1206	IPv6 Fragmentation:
1207	
1208	ip6frag_high_thresh - INTEGER
1209		Maximum memory used to reassemble IPv6 fragments. When
1210		ip6frag_high_thresh bytes of memory is allocated for this purpose,
1211		the fragment handler will toss packets until ip6frag_low_thresh
1212		is reached.
1213	
1214	ip6frag_low_thresh - INTEGER
1215		See ip6frag_high_thresh
1216	
1217	ip6frag_time - INTEGER
1218		Time in seconds to keep an IPv6 fragment in memory.
1219	
1220	conf/default/*:
1221		Change the interface-specific default settings.
1222	
1223	
1224	conf/all/*:
1225		Change all the interface-specific settings.
1226	
1227		[XXX:  Other special features than forwarding?]
1228	
1229	conf/all/forwarding - BOOLEAN
1230		Enable global IPv6 forwarding between all interfaces.
1231	
1232		IPv4 and IPv6 work differently here; e.g. netfilter must be used
1233		to control which interfaces may forward packets and which not.
1234	
1235		This also sets all interfaces' Host/Router setting
1236		'forwarding' to the specified value.  See below for details.
1237	
1238		This referred to as global forwarding.
1239	
1240	proxy_ndp - BOOLEAN
1241		Do proxy ndp.
1242	
1243	fwmark_reflect - BOOLEAN
1244		Controls the fwmark of kernel-generated IPv6 reply packets that are not
1245		associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
1246		If unset, these packets have a fwmark of zero. If set, they have the
1247		fwmark of the packet they are replying to.
1248		Default: 0
1249	
1250	conf/interface/*:
1251		Change special settings per interface.
1252	
1253		The functional behaviour for certain settings is different
1254		depending on whether local forwarding is enabled or not.
1255	
1256	accept_ra - INTEGER
1257		Accept Router Advertisements; autoconfigure using them.
1258	
1259		It also determines whether or not to transmit Router
1260		Solicitations. If and only if the functional setting is to
1261		accept Router Advertisements, Router Solicitations will be
1262		transmitted.
1263	
1264		Possible values are:
1265			0 Do not accept Router Advertisements.
1266			1 Accept Router Advertisements if forwarding is disabled.
1267			2 Overrule forwarding behaviour. Accept Router Advertisements
1268			  even if forwarding is enabled.
1269	
1270		Functional default: enabled if local forwarding is disabled.
1271				    disabled if local forwarding is enabled.
1272	
1273	accept_ra_defrtr - BOOLEAN
1274		Learn default router in Router Advertisement.
1275	
1276		Functional default: enabled if accept_ra is enabled.
1277				    disabled if accept_ra is disabled.
1278	
1279	accept_ra_from_local - BOOLEAN
1280		Accept RA with source-address that is found on local machine
1281	        if the RA is otherwise proper and able to be accepted.
1282	        Default is to NOT accept these as it may be an un-intended
1283	        network loop.
1284	
1285		Functional default:
1286	           enabled if accept_ra_from_local is enabled
1287	               on a specific interface.
1288		   disabled if accept_ra_from_local is disabled
1289	               on a specific interface.
1290	
1291	accept_ra_pinfo - BOOLEAN
1292		Learn Prefix Information in Router Advertisement.
1293	
1294		Functional default: enabled if accept_ra is enabled.
1295				    disabled if accept_ra is disabled.
1296	
1297	accept_ra_rt_info_max_plen - INTEGER
1298		Maximum prefix length of Route Information in RA.
1299	
1300		Route Information w/ prefix larger than or equal to this
1301		variable shall be ignored.
1302	
1303		Functional default: 0 if accept_ra_rtr_pref is enabled.
1304				    -1 if accept_ra_rtr_pref is disabled.
1305	
1306	accept_ra_rtr_pref - BOOLEAN
1307		Accept Router Preference in RA.
1308	
1309		Functional default: enabled if accept_ra is enabled.
1310				    disabled if accept_ra is disabled.
1311	
1312	accept_ra_mtu - BOOLEAN
1313		Apply the MTU value specified in RA option 5 (RFC4861). If
1314		disabled, the MTU specified in the RA will be ignored.
1315	
1316		Functional default: enabled if accept_ra is enabled.
1317				    disabled if accept_ra is disabled.
1318	
1319	accept_redirects - BOOLEAN
1320		Accept Redirects.
1321	
1322		Functional default: enabled if local forwarding is disabled.
1323				    disabled if local forwarding is enabled.
1324	
1325	accept_source_route - INTEGER
1326		Accept source routing (routing extension header).
1327	
1328		>= 0: Accept only routing header type 2.
1329		< 0: Do not accept routing header.
1330	
1331		Default: 0
1332	
1333	autoconf - BOOLEAN
1334		Autoconfigure addresses using Prefix Information in Router
1335		Advertisements.
1336	
1337		Functional default: enabled if accept_ra_pinfo is enabled.
1338				    disabled if accept_ra_pinfo is disabled.
1339	
1340	dad_transmits - INTEGER
1341		The amount of Duplicate Address Detection probes to send.
1342		Default: 1
1343	
1344	forwarding - INTEGER
1345		Configure interface-specific Host/Router behaviour.
1346	
1347		Note: It is recommended to have the same setting on all
1348		interfaces; mixed router/host scenarios are rather uncommon.
1349	
1350		Possible values are:
1351			0 Forwarding disabled
1352			1 Forwarding enabled
1353	
1354		FALSE (0):
1355	
1356		By default, Host behaviour is assumed.  This means:
1357	
1358		1. IsRouter flag is not set in Neighbour Advertisements.
1359		2. If accept_ra is TRUE (default), transmit Router
1360		   Solicitations.
1361		3. If accept_ra is TRUE (default), accept Router
1362		   Advertisements (and do autoconfiguration).
1363		4. If accept_redirects is TRUE (default), accept Redirects.
1364	
1365		TRUE (1):
1366	
1367		If local forwarding is enabled, Router behaviour is assumed.
1368		This means exactly the reverse from the above:
1369	
1370		1. IsRouter flag is set in Neighbour Advertisements.
1371		2. Router Solicitations are not sent unless accept_ra is 2.
1372		3. Router Advertisements are ignored unless accept_ra is 2.
1373		4. Redirects are ignored.
1374	
1375		Default: 0 (disabled) if global forwarding is disabled (default),
1376			 otherwise 1 (enabled).
1377	
1378	hop_limit - INTEGER
1379		Default Hop Limit to set.
1380		Default: 64
1381	
1382	mtu - INTEGER
1383		Default Maximum Transfer Unit
1384		Default: 1280 (IPv6 required minimum)
1385	
1386	router_probe_interval - INTEGER
1387		Minimum interval (in seconds) between Router Probing described
1388		in RFC4191.
1389	
1390		Default: 60
1391	
1392	router_solicitation_delay - INTEGER
1393		Number of seconds to wait after interface is brought up
1394		before sending Router Solicitations.
1395		Default: 1
1396	
1397	router_solicitation_interval - INTEGER
1398		Number of seconds to wait between Router Solicitations.
1399		Default: 4
1400	
1401	router_solicitations - INTEGER
1402		Number of Router Solicitations to send until assuming no
1403		routers are present.
1404		Default: 3
1405	
1406	use_tempaddr - INTEGER
1407		Preference for Privacy Extensions (RFC3041).
1408		  <= 0 : disable Privacy Extensions
1409		  == 1 : enable Privacy Extensions, but prefer public
1410		         addresses over temporary addresses.
1411		  >  1 : enable Privacy Extensions and prefer temporary
1412		         addresses over public addresses.
1413		Default:  0 (for most devices)
1414			 -1 (for point-to-point devices and loopback devices)
1415	
1416	temp_valid_lft - INTEGER
1417		valid lifetime (in seconds) for temporary addresses.
1418		Default: 604800 (7 days)
1419	
1420	temp_prefered_lft - INTEGER
1421		Preferred lifetime (in seconds) for temporary addresses.
1422		Default: 86400 (1 day)
1423	
1424	max_desync_factor - INTEGER
1425		Maximum value for DESYNC_FACTOR, which is a random value
1426		that ensures that clients don't synchronize with each
1427		other and generate new addresses at exactly the same time.
1428		value is in seconds.
1429		Default: 600
1430	
1431	regen_max_retry - INTEGER
1432		Number of attempts before give up attempting to generate
1433		valid temporary addresses.
1434		Default: 5
1435	
1436	max_addresses - INTEGER
1437		Maximum number of autoconfigured addresses per interface.  Setting
1438		to zero disables the limitation.  It is not recommended to set this
1439		value too large (or to zero) because it would be an easy way to
1440		crash the kernel by allowing too many addresses to be created.
1441		Default: 16
1442	
1443	disable_ipv6 - BOOLEAN
1444		Disable IPv6 operation.  If accept_dad is set to 2, this value
1445		will be dynamically set to TRUE if DAD fails for the link-local
1446		address.
1447		Default: FALSE (enable IPv6 operation)
1448	
1449		When this value is changed from 1 to 0 (IPv6 is being enabled),
1450		it will dynamically create a link-local address on the given
1451		interface and start Duplicate Address Detection, if necessary.
1452	
1453		When this value is changed from 0 to 1 (IPv6 is being disabled),
1454		it will dynamically delete all address on the given interface.
1455	
1456	accept_dad - INTEGER
1457		Whether to accept DAD (Duplicate Address Detection).
1458		0: Disable DAD
1459		1: Enable DAD (default)
1460		2: Enable DAD, and disable IPv6 operation if MAC-based duplicate
1461		   link-local address has been found.
1462	
1463	force_tllao - BOOLEAN
1464		Enable sending the target link-layer address option even when
1465		responding to a unicast neighbor solicitation.
1466		Default: FALSE
1467	
1468		Quoting from RFC 2461, section 4.4, Target link-layer address:
1469	
1470		"The option MUST be included for multicast solicitations in order to
1471		avoid infinite Neighbor Solicitation "recursion" when the peer node
1472		does not have a cache entry to return a Neighbor Advertisements
1473		message.  When responding to unicast solicitations, the option can be
1474		omitted since the sender of the solicitation has the correct link-
1475		layer address; otherwise it would not have be able to send the unicast
1476		solicitation in the first place. However, including the link-layer
1477		address in this case adds little overhead and eliminates a potential
1478		race condition where the sender deletes the cached link-layer address
1479		prior to receiving a response to a previous solicitation."
1480	
1481	ndisc_notify - BOOLEAN
1482		Define mode for notification of address and device changes.
1483		0 - (default): do nothing
1484		1 - Generate unsolicited neighbour advertisements when device is brought
1485		    up or hardware address changes.
1486	
1487	mldv1_unsolicited_report_interval - INTEGER
1488		The interval in milliseconds in which the next unsolicited
1489		MLDv1 report retransmit will take place.
1490		Default: 10000 (10 seconds)
1491	
1492	mldv2_unsolicited_report_interval - INTEGER
1493		The interval in milliseconds in which the next unsolicited
1494		MLDv2 report retransmit will take place.
1495		Default: 1000 (1 second)
1496	
1497	force_mld_version - INTEGER
1498		0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
1499		1 - Enforce to use MLD version 1
1500		2 - Enforce to use MLD version 2
1501	
1502	suppress_frag_ndisc - INTEGER
1503		Control RFC 6980 (Security Implications of IPv6 Fragmentation
1504		with IPv6 Neighbor Discovery) behavior:
1505		1 - (default) discard fragmented neighbor discovery packets
1506		0 - allow fragmented neighbor discovery packets
1507	
1508	optimistic_dad - BOOLEAN
1509		Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
1510			0: disabled (default)
1511			1: enabled
1512	
1513	use_optimistic - BOOLEAN
1514		If enabled, do not classify optimistic addresses as deprecated during
1515		source address selection.  Preferred addresses will still be chosen
1516		before optimistic addresses, subject to other ranking in the source
1517		address selection algorithm.
1518			0: disabled (default)
1519			1: enabled
1520	
1521	icmp/*:
1522	ratelimit - INTEGER
1523		Limit the maximal rates for sending ICMPv6 packets.
1524		0 to disable any limiting,
1525		otherwise the minimal space between responses in milliseconds.
1526		Default: 1000
1527	
1528	
1529	IPv6 Update by:
1530	Pekka Savola <pekkas@netcore.fi>
1531	YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1532	
1533	
1534	/proc/sys/net/bridge/* Variables:
1535	
1536	bridge-nf-call-arptables - BOOLEAN
1537		1 : pass bridged ARP traffic to arptables' FORWARD chain.
1538		0 : disable this.
1539		Default: 1
1540	
1541	bridge-nf-call-iptables - BOOLEAN
1542		1 : pass bridged IPv4 traffic to iptables' chains.
1543		0 : disable this.
1544		Default: 1
1545	
1546	bridge-nf-call-ip6tables - BOOLEAN
1547		1 : pass bridged IPv6 traffic to ip6tables' chains.
1548		0 : disable this.
1549		Default: 1
1550	
1551	bridge-nf-filter-vlan-tagged - BOOLEAN
1552		1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1553		0 : disable this.
1554		Default: 0
1555	
1556	bridge-nf-filter-pppoe-tagged - BOOLEAN
1557		1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1558		0 : disable this.
1559		Default: 0
1560	
1561	bridge-nf-pass-vlan-input-dev - BOOLEAN
1562		1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
1563		interface on the bridge and set the netfilter input device to the vlan.
1564		This allows use of e.g. "iptables -i br0.1" and makes the REDIRECT
1565		target work with vlan-on-top-of-bridge interfaces.  When no matching
1566		vlan interface is found, or this switch is off, the input device is
1567		set to the bridge interface.
1568		0: disable bridge netfilter vlan interface lookup.
1569		Default: 0
1570	
1571	proc/sys/net/sctp/* Variables:
1572	
1573	addip_enable - BOOLEAN
1574		Enable or disable extension of  Dynamic Address Reconfiguration
1575		(ADD-IP) functionality specified in RFC5061.  This extension provides
1576		the ability to dynamically add and remove new addresses for the SCTP
1577		associations.
1578	
1579		1: Enable extension.
1580	
1581		0: Disable extension.
1582	
1583		Default: 0
1584	
1585	addip_noauth_enable - BOOLEAN
1586		Dynamic Address Reconfiguration (ADD-IP) requires the use of
1587		authentication to protect the operations of adding or removing new
1588		addresses.  This requirement is mandated so that unauthorized hosts
1589		would not be able to hijack associations.  However, older
1590		implementations may not have implemented this requirement while
1591		allowing the ADD-IP extension.  For reasons of interoperability,
1592		we provide this variable to control the enforcement of the
1593		authentication requirement.
1594	
1595		1: Allow ADD-IP extension to be used without authentication.  This
1596		   should only be set in a closed environment for interoperability
1597		   with older implementations.
1598	
1599		0: Enforce the authentication requirement
1600	
1601		Default: 0
1602	
1603	auth_enable - BOOLEAN
1604		Enable or disable Authenticated Chunks extension.  This extension
1605		provides the ability to send and receive authenticated chunks and is
1606		required for secure operation of Dynamic Address Reconfiguration
1607		(ADD-IP) extension.
1608	
1609		1: Enable this extension.
1610		0: Disable this extension.
1611	
1612		Default: 0
1613	
1614	prsctp_enable - BOOLEAN
1615		Enable or disable the Partial Reliability extension (RFC3758) which
1616		is used to notify peers that a given DATA should no longer be expected.
1617	
1618		1: Enable extension
1619		0: Disable
1620	
1621		Default: 1
1622	
1623	max_burst - INTEGER
1624		The limit of the number of new packets that can be initially sent.  It
1625		controls how bursty the generated traffic can be.
1626	
1627		Default: 4
1628	
1629	association_max_retrans - INTEGER
1630		Set the maximum number for retransmissions that an association can
1631		attempt deciding that the remote end is unreachable.  If this value
1632		is exceeded, the association is terminated.
1633	
1634		Default: 10
1635	
1636	max_init_retransmits - INTEGER
1637		The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
1638		that an association will attempt before declaring the destination
1639		unreachable and terminating.
1640	
1641		Default: 8
1642	
1643	path_max_retrans - INTEGER
1644		The maximum number of retransmissions that will be attempted on a given
1645		path.  Once this threshold is exceeded, the path is considered
1646		unreachable, and new traffic will use a different path when the
1647		association is multihomed.
1648	
1649		Default: 5
1650	
1651	pf_retrans - INTEGER
1652		The number of retransmissions that will be attempted on a given path
1653		before traffic is redirected to an alternate transport (should one
1654		exist).  Note this is distinct from path_max_retrans, as a path that
1655		passes the pf_retrans threshold can still be used.  Its only
1656		deprioritized when a transmission path is selected by the stack.  This
1657		setting is primarily used to enable fast failover mechanisms without
1658		having to reduce path_max_retrans to a very low value.  See:
1659		http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
1660		for details.  Note also that a value of pf_retrans > path_max_retrans
1661		disables this feature
1662	
1663		Default: 0
1664	
1665	rto_initial - INTEGER
1666		The initial round trip timeout value in milliseconds that will be used
1667		in calculating round trip times.  This is the initial time interval
1668		for retransmissions.
1669	
1670		Default: 3000
1671	
1672	rto_max - INTEGER
1673		The maximum value (in milliseconds) of the round trip timeout.  This
1674		is the largest time interval that can elapse between retransmissions.
1675	
1676		Default: 60000
1677	
1678	rto_min - INTEGER
1679		The minimum value (in milliseconds) of the round trip timeout.  This
1680		is the smallest time interval the can elapse between retransmissions.
1681	
1682		Default: 1000
1683	
1684	hb_interval - INTEGER
1685		The interval (in milliseconds) between HEARTBEAT chunks.  These chunks
1686		are sent at the specified interval on idle paths to probe the state of
1687		a given path between 2 associations.
1688	
1689		Default: 30000
1690	
1691	sack_timeout - INTEGER
1692		The amount of time (in milliseconds) that the implementation will wait
1693		to send a SACK.
1694	
1695		Default: 200
1696	
1697	valid_cookie_life - INTEGER
1698		The default lifetime of the SCTP cookie (in milliseconds).  The cookie
1699		is used during association establishment.
1700	
1701		Default: 60000
1702	
1703	cookie_preserve_enable - BOOLEAN
1704		Enable or disable the ability to extend the lifetime of the SCTP cookie
1705		that is used during the establishment phase of SCTP association
1706	
1707		1: Enable cookie lifetime extension.
1708		0: Disable
1709	
1710		Default: 1
1711	
1712	cookie_hmac_alg - STRING
1713		Select the hmac algorithm used when generating the cookie value sent by
1714		a listening sctp socket to a connecting client in the INIT-ACK chunk.
1715		Valid values are:
1716		* md5
1717		* sha1
1718		* none
1719		Ability to assign md5 or sha1 as the selected alg is predicated on the
1720		configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and
1721		CONFIG_CRYPTO_SHA1).
1722	
1723		Default: Dependent on configuration.  MD5 if available, else SHA1 if
1724		available, else none.
1725	
1726	rcvbuf_policy - INTEGER
1727		Determines if the receive buffer is attributed to the socket or to
1728		association.   SCTP supports the capability to create multiple
1729		associations on a single socket.  When using this capability, it is
1730		possible that a single stalled association that's buffering a lot
1731		of data may block other associations from delivering their data by
1732		consuming all of the receive buffer space.  To work around this,
1733		the rcvbuf_policy could be set to attribute the receiver buffer space
1734		to each association instead of the socket.  This prevents the described
1735		blocking.
1736	
1737		1: rcvbuf space is per association
1738		0: rcvbuf space is per socket
1739	
1740		Default: 0
1741	
1742	sndbuf_policy - INTEGER
1743		Similar to rcvbuf_policy above, this applies to send buffer space.
1744	
1745		1: Send buffer is tracked per association
1746		0: Send buffer is tracked per socket.
1747	
1748		Default: 0
1749	
1750	sctp_mem - vector of 3 INTEGERs: min, pressure, max
1751		Number of pages allowed for queueing by all SCTP sockets.
1752	
1753		min: Below this number of pages SCTP is not bothered about its
1754		memory appetite. When amount of memory allocated by SCTP exceeds
1755		this number, SCTP starts to moderate memory usage.
1756	
1757		pressure: This value was introduced to follow format of tcp_mem.
1758	
1759		max: Number of pages allowed for queueing by all SCTP sockets.
1760	
1761		Default is calculated at boot time from amount of available memory.
1762	
1763	sctp_rmem - vector of 3 INTEGERs: min, default, max
1764		Only the first value ("min") is used, "default" and "max" are
1765		ignored.
1766	
1767		min: Minimal size of receive buffer used by SCTP socket.
1768		It is guaranteed to each SCTP socket (but not association) even
1769		under moderate memory pressure.
1770	
1771		Default: 1 page
1772	
1773	sctp_wmem  - vector of 3 INTEGERs: min, default, max
1774		Currently this tunable has no effect.
1775	
1776	addr_scope_policy - INTEGER
1777		Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
1778	
1779		0   - Disable IPv4 address scoping
1780		1   - Enable IPv4 address scoping
1781		2   - Follow draft but allow IPv4 private addresses
1782		3   - Follow draft but allow IPv4 link local addresses
1783	
1784		Default: 1
1785	
1786	
1787	/proc/sys/net/core/*
1788		Please see: Documentation/sysctl/net.txt for descriptions of these entries.
1789	
1790	
1791	/proc/sys/net/unix/*
1792	max_dgram_qlen - INTEGER
1793		The maximum length of dgram socket receive queue
1794	
1795		Default: 10
1796	
1797	
1798	UNDOCUMENTED:
1799	
1800	/proc/sys/net/irda/*
1801		fast_poll_increase FIXME
1802		warn_noreply_time FIXME
1803		discovery_slots FIXME
1804		slot_timeout FIXME
1805		max_baud_rate FIXME
1806		discovery_timeout FIXME
1807		lap_keepalive_time FIXME
1808		max_noreply_time FIXME
1809		max_tx_data_size FIXME
1810		max_tx_window FIXME
1811		min_tx_turn_time FIXME
Hide Line Numbers
About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Information is copyright its respective author. All material is available from the Linux Kernel Source distributed under a GPL License. This page is provided as a free service by mjmwired.net.